Yashica ML 28-85 3.5-4.5

Some nice photos, and showing yet another reason not to bypass a lens for it's perceived issues that some lenses (sadly) get a bad rep for. I also think you're illustrating what might be part of a dying ritual of an older and a more methodical era, where one actually takes the time to test a lens under a variety of conditions, instead of assuming that "my well researched lens is perfectly capable of performing brilliantly under any and every circumstance".

You get a couple of added bonuses to using it on a M4/3 body too... cropping to the sweet spot, and a tighter field of view when capturing any 'unwilling to pose' fauna that are usually hard to get close to. When hunger takes over for the fear of people, it's gets easier, but it helps to tighten up for better details when you get the built-in byproduct of cropped sensors using full frame legacy glass.

As far as adapters go, I've had my own issues with them and have relearned the same lesson over and over again - that acceptable quality should never be assumed, and it also isn't a regulated requirement where everyone will get XX amount of it with every purchase. I've pretty much limited the generic adapters I use, to K&F (Sony) and FotoDiox (Sony and M4/3). The main adapter I use is a Novoflex C/Y to E-Mount, and it was well worth the winning bid at auction to get one. The cheap Chinese adapter quality can spin on a dime to be wildly out of whack more often than you'd think should be possible.

You bring up the point about differences across formats, but most users might tend to forget that in the digital world, there's no such thing as a universal sensor. In the film side of adaption, everyone has an additional choice of what film stock to use. Not every roll of color negative is the same in color rendition and it's obvious that when swapped out with the same 'hardware', there will be a different result seen in the finished photo. The same applies to sensors, but many users assume (maybe it's because they're led to it by the maker) that sensors are neutral. They are tuned in somewhat the same way that film is.

It even happens when the underlying camera is rebranded. A Leitz V-Lux (any version) is a rebranded Panasonic Lumix, and other than logo changes on the outside and a substitution of glass - there's a reworked sensor to match Leica's preferences. It also happens when each maker designs and programs the purchased sensor's processor chips that they buy from a select few sensor makers. Sony sensors are in far more cameras than those branded under the Sony label, as not every camera maker makes their own sensors. Samsung and Toshiba have their sensors in multiple cameras, just like Sony does.

DxO started their company life as an engineering firm that helped camera makers with their digital 'tuning', and sometime later they started to produce their own digital darkroom software. As part of that software's functionality, I can in effect, retune my Sony Exmor sensors to render like I'd used a Hasselblad X1D or a Nikon Z6 body. The differences can sometimes be subtle, but obvious when you can quickly swap their digital profiles to view the results on-screen.