once again, the "having video makes this camera worse for stills" argument
I completely disagree
I do not think including video itself makes it 'worse', but pointing out that video is not just a 'free' feature. It eats development time, it eats limited resources on the camera, it eats testing time, it effects design decisions the impact the still photo side. To a limited degree, this is a zero sum case and, if the body did not include video, the same amount of resources being put into it likely would have produced a better still camera.
Naturally it could be argued that the increase in sales/demand for a dual purpose camera out weighs this and TBH I have no idea if it does or not.
and I didn't hear any stills shooters complaining in the times of film, when you could buy better and vastly cheaper film stocks thanks to the huge demand for film by the movie industry
just as "kodak should forget about stills and develop better film stocks for its main customer, the movie industry" was a dumb argument back then, "canon should forget about video and develop better stills cameras" is silly now
Not really the same thing. I do not think anyone is suggesting Canon should not product video cameras, or even dual function cameras, just that Canon should also produce cameras where the only function is still and thus all design decisions and hardware/firmware tradeoffs focus on still photography rather then 'it has to work well for both'.
When you buy a dual use camera, you are paying for not only the increased manufacturing cost of including video capabilities and the R&D/Testing involved in a more complex device, but, as said, if this theory is true, you are paying for design decisions intended to benefit a whole class of use-cases that you (assuming one only wants stills) do not care about, and thus design decisions that focus on the other use cases probably would have designed it differently.
For reference, I spent years working in embedded systems (not cameras) and this was always a major point of contention, esp when marketing pushed for features that a certain percentage of customers wanted to be put into all models, even when those customers were already served by specialized models that met their use cases specifically. It was very frustrating from an engineering perspective.