Possibly one of the most polarising food items on this planet: Marmite. It seems you have to be English to be able to like it, but it used to be widely available in mainland Europe as well. Nowadays, not as much, as far as I can see. Perhaps a Brexit side effect. In any case, some will lament its poorer availability, and others will think “good riddance”. Marmite doesn’t do mid-field, only extremes.
I’m sure I ate it as a kid, and if my memory doesn’t fail me, I didn’t hate it. But many years passed since, and I have no idea how I’d react now. My taste buds have changed enough over time, so no way to tell without trying again.

This isn’t a food blog anyway. But Marmite works well as a generic description of products that only seem to evoke extreme reactions. Leica’s products tend to fall in that category, even if that also seems a case of rather extreme fanboy-ism.
No, more specifically I think about a number of lenses, across many brands. Lenses that some feel are utterly hopeless, and others feel they’re utterly glorious.
Perfection and Measurable Performance
Those serious about their photography will sooner or later spend serious money to find a lens for the lens mount of choice. Making that choice can be pretty hard. For most lens mounts, there is plenty choice, and each and every available product makes a different set of compromises.
The perfect 15-1000mm f/1.4 lens doesn’t exist. If it would exist, it would probably be wildly expensive, insanely large and heavy. That basically already sets the simplest set of constraints: price, weight, size.
Perfect optical performance is equally rare. There are a number of potential issues which will affect the final image. Vignetting and purple fringing (more scientifically chromatic aberration) are relatively known, but there are more considerations such as flatness of field, longitudinal chromatic aberrations, coma, flare. And the most known one: sharpness. As the definition of sharpness is a subject in itself, let’s split that up into resolving power (how much detail in the lens capable of transmitting) and contrast.

All the above points are measurable. And that is what reviews of lenses tend to present. As such, a lens that measures excellent across these points could be thought of as approaching perfection. A worthy addition to your gear collection.
Imperfection and Unmeasurables
But are they? Is measurable perfection the goal? This is where things go totally marmite.
Part of the problem is that there are some areas that cannot be measured, and that the way you use a lens may or may not emphasise its strong points or weak points. That of course shifts the question of value a lot.
This last bit sounds a bit abstract, so a simple example. Several years ago, I owned a Nikkor AF-D 80-200mm f/2.8 – a substantial telezoom with large aperture. This lens had generally a very high performance, but one serious optical weak point: at close focussing distances with wide apertures, it simply wasn’t very sharp. For my uses, this didn’t really matter since I wasn’t using the lens that way. Would it be your bread-and-butter lens for portraits, you’d have a pretty serious issue. For me, the value of the lens was fine. For a portrait photographer, it would be far from fine. The second issue was that it’s a large, heavy lens, and that was the issue for me: I didn’t use it enough to justify carrying it around. The 105mm f/2.5 typically covered it for me.

In short, the point is that value isn’t only about cost, but also about what you get out of it.
But the real marmite stuff is the unmeasurable characteristics of a lens. The most popular term there is bokeh. A term I increasingly dislike, since it’s mostly used wrong, but it’s one to look at. The other ones are micro-contrast and 3D pop. All completely undefined, unmeasurable things. Yet, people see them.
Bokeh and its brothers
Somehow for many, bokeh has grown to mean the effect of having large, out of focus diffuse balls of light in the background of an image. That is not what bokeh is, though. It actually is the aesthetic quality of the out-of-focus areas. So, it is not a quantity (meaning, you cannot have a lot of bokeh, or little), and it’s not an effect of sorts. It is how a lens renders the out-of-focus area. Being an aesthetic quality, it means it comes down to taste. Most like it to be subtle, calming, and the out of focus areas shouldn’t distract from what is in focus. Considering this, I think the shot above shows that the bokeh of the 105mm f/2.5 lens is very good. And I think most would agree.
But it can take different shapes and forms. Ultimately, it is about what works for you. Your personal preference applies. If you always shoot at small apertures, minimising your out-of-focus areas, the whole point of bokeh is moot anyway.

Aside from how the out of focus area looks, the transition from in focus to out of focus also matters. It’s not uncommon to see some irregularities there that make the image seem more ‘nervous’, rather than a gradual drop-off.
On bokeh, there seems to be a level of agreement what it means, and how much it can affect an image. The other characteristics I mentioned before (micro contrast and 3D pop) are much less defined. And honestly, I’m not going to bother too much. Micro contrast seems related to the crispness of contrast transitions (light to dark) in detailed areas, and 3D pop seems to be about subject separation (so related to the transition areas mentioned above). Not entirely sure on these two points, as they’re not specifically things I tend to look for.
Characteristics, or character?
Even if bokeh and its brothers are not measurable, they’re still identifiable – images from different lenses can look quite different. The better lens tests do attempt to show it and describe it. I think that is positive; even if not measurable, it’s worth reporting on. Pure lab tests (shots of resolution charts and grids and such) won’t show these differences.
The next step in the marmite mess is the overall effect of it all. Proper lens reviews will show this. Example images, and a real photographers’ view on the specifics of the results you can expect for a lens. Real photos, taken in different realistic situations exploring the results that can be expected.
But this is also typically where the bickering and disagreements start. It opens the door on lenses with poor test result, but the resulting images showing something special. Those that favour lab tests seem to overlook that most of us aren’t shooting test charts to measure our equipment. We’re shooting three dimensional subjects, in varying light conditions. How that works out matters, even if it’s difficult to measure and put in numbers.
Still, how far should that go? Does a lens expose its character, or better put: does a lens leave its imprint on the final image? And if yes, how desirable is that?
Many will say it is not desirable. If you’d want that specific character to your image, you can always add it in post processing. And that’s true, and with software getting more advanced increasingly true. What is also true is that it means a part of “making the intended image” is moved from during the shoot, to after the shoot. Whether that is preferred is again a personal choice.
For those who prefer to reduce post-processing, getting the lens that renders the way you like is worth it. That might still be that perfect lens, that measurably scores perfect. But it could also be that lens with flaws, where the combination of imperfections just render results that are exactly what you hope for.

Reality is, every single lens makes a compromise somewhere. The validity of that compromise comes down to your expectations, and what you’re willing to accept.
Point is that going simply by the numbers and choosing the lens that tests as the best, may not always be the way to go. Lab tests sure have their use, but sample images and notes from actual users are at least as important.
My choice
If people ask me for a recommendation, I will generally point to the lenses that do well in lab tests, but noting that it’s worth checking samples images and user reviews too. Blindly recommending to buy a couple of (heavy and expensive) f/2.8 zooms “because they’re the best” or using horrible nicknames like Holy Trinity of Lenses, not my thing. Yes, they’re very good lenses, but that doesn’t make them right for all.
For my own uses, I tend to prefer the lenses that aren’t perfect. Poor lab testing result, clear and obvious optical flaws. Poor scores. And yet, the results… just work. For me.
There is more to it; how does it feel to use, is it nicely balanced on my camera? Weight, price and size factor in.
I have some lenses where there is (I think) general consensus that they are very good; two passed here as examples (Nikkor 28mm f/1.4E and the classic 105mm f/2.5). They work very well for me. Another optical marvel, which I don’t have anymore, is the Elmarit-R 60mm macro above. As good as it may be, I never bonded with it. Terribly long focus throw, and missing something in its character – didn’t like that lens much.
There are other lenses I like much more that aren’t all that brilliant. I like them because of, or thanks to, their flaws. They’re complete marmite lenses. I would never recommend anyone else to buy one of those, though – they’re right for me.

I guess my only point here is that it doesn’t matter whether you like marmite or not. If you don’t, fine. Don’t buy it, and leave it at that. Don’t tell others it’s nasty, because they might have different taste. If you do like it, eat it all you want. Just don’t convince others to do the same.
The idea that an lens who tests best in lab conditions is the best lens period, is a fallacy. The best lens is one that fits your uses, gets you images that make you smile and fits your budget. And from there on, use it till it falls apart.
P.S. (1) The choice for car images is not entirely without reason. Modern cars are getting more and more perfect – in economy, noise, efficiency, performance. But the question whether they’re more fun to drive than an older car with quirks, a manual gearbox and oily maintenance…. take your pick. There is no right or wrong. But I’ll take that marmite manual gearbox, thank you.
P.S. (2) As mentioned in previous post, the ongoing pandemic doesn’t help my creativity, so instead I’ll start rambling about my gear. And lenses are the critical part of the puzzle for me. Perhaps I’ll figure out which gear I actually really need, and manage to get the collection down to normal proportions. Who knows.