My POV comes from having helped hundreds of people buy new gear, listening to their issues and concerns, and then helping them find the best "fit" for their needs.
It's been my experience that by the time people reach the "good photographer" level, that they are almost universally using what would be considered more or less state of the industry type gear. The Reuters survey of what lenses were used in 2013 pretty much proved that point; the Canon 16-35 f/2.8 L lens was used by more people for more photos than basically, all other lenses combined, around the world. This has been this way for decades, and that's why I agreed, the four genralizations you started with as a premise DO HAVE at least some basis in reality.
Here's one of the Reuters survey results articles. Lens-wise, 52.9% of the images were shot with the Canon 16-35/2.8-L zoom. Next was the 70-200 with 38.2%. The 24-105, and 24-70, and 100-400mm lenses each had TINY slices of the reminder; yes....the 24-70 was used for only a small fraction of Reuters news photos! f/2.8 was the most commonly-used aperture.
I don't think there's selection bias, because the survey is very clear on what its population is. It's working photojournalists that submit to Reuters and other major outlets. If I'm releasing a survey of what music 18-24 year olds like, the survey isn't biased because I left out 30 year olds. It's simply a survey of a different demographic. Same thing here. Just because they left out landscape photographers, doesn't make it biased, it makes it a different survey.
These types are going to settle in on a couple of workhorse lenses because they're easy to get a replacement if your is stolen, sent in for repair or lost by the airline. Doing sports professionally, the pros only use a very small handful of lenses: basically Canon and Nikon's f/2.8 70-200 and their f/2.8 300s.
I don't think the survey set out to make any claims about other fields of photography.
I don't think the survey is biased, those two lenses are a photojournalist's goto, everyday, workhorses. It isn't about replacement lenses, a staff photographer has access to practically any lens they desire. It is about what works for them, what lens enable them to easily get the shot the want. I suspect that the majority of images were shot at the extremes, 16mm, 35mm, 70mm and 200m (or close to the extremes). I tend to use a single zoom as two primes, I suspect most photojournalists shoot similarly.
Since Derrel bases a lot of his opinion on his background selling stuff to people, let me say that I base my opinion on years of teaching stats and experimental design to various grad student disciplines.
Although the survey limits itself to the PJs who submit to Reuters, Derrel didn't use those criteria when quoting it and that is the exact danger for that kind of survey.
There are many reasons why working PJs could be using those lenses and bodies - many are probably hidden and probably very few apply to any group outside of high level working PJs.
Maybe the PJs got better deals on Canon stuff.
Maybe the experienced PJs had a huge inventory of Canon stuff
Maybe Canon Professional services gave better service.
Maybe they liked the feel of the bodies or lenses
Maybe those lenses are more ubiquitous.
Maybe editors like the FOV of those kinds of images.
Who knows, but quoting the survey without any qualifiers makes believe that selection bias doesn't exist.
Without knowing the effects of the sampling then there is no way to know if the results can be extended outside the sample.