I'm not sure I follow all philosophical turns in the discussion, but anyway some more loose talk:
in the film days the post-processing possibilities were limited, so if we wanted to make a landscape image really pop we loaded Velvia 50, which of course was not designed to be accurate but to make landscape scenes pop. I guess color couldn't be 100% designed back then either, the chemicals reacted in certain ways so you had to take the good with the bad.
With digital we have total control over color and can design it. However, photographers buying into digital came from film once and they didn't all of a sudden become their own post-processing experts so in the start profiles had to provide looks much in the same way as film did.
However, today things are different. Photographers are more used to post-processing, the software is better at it so you can design your own look quite easily (although I think there is room for improvement still!). Why should then the camera profile apply some preset film-like look? What is the value to the photographer to get a rendering with a different hue than the eye saw? (Assuming normal light conditions)
Using Hasseblad as an example again, I'd say that their look is 95% realistic. If you shoot a scene with reasonable light, look at the default render you say "wow it looks exactly the same", if the scene was dull, the result is dull. It's only when you start to look really closely some smaller deviations can be found. If you would look at say a Velvia 50 shot you'd spot all sorts of hue errors and certainly way too high saturation, but it can make a dull scene pop. That is I'd say Hasselblad has to a large extent followed the principle that the camera should first and foremost deliver a realistic image, and the creative look should be left to the photographer. Then there are other manufacturers that have stayed with the tradition of choosing a film stock for creative purposes, or like Adobe provide the one true film look(tm) that we're supposed to like.
Another interesting aspect of Hasseblad is that they have made their different models look almost exactly the same, even if there's a Kodak sensor in one model and a Dalsa in another. I've heard many professional photographers appreciate this feature, as they can change camera in the middle of a shoot, or use different camera models simultaneously during a shoot. Then there are other manufacturers that make their models look differently. The reputation that Kodak has a saturated look and Dalsa better skin tones doesn't come from Hasseblad's products, but from Phase One's, mainly P45+ (Kodak) vs P65+ (Dalsa) I suppose. There are hardware differences, but they don't need to result in very large color differences if the profiles are designed for the same targeted look.
(It should be said that Hassy's look and all other profiles I've tested have quite large errors on high saturation colors, but I think that is more about gamut mapping making images easier to print, or simply because high saturation colors haven't been prioritized when designing the profile)
I do not fully agree that the talk about that raw conversion and tone curves etc doesn't mean anything. I does. Just maybe not as much as we are lead to believe. Say if all manufacturers would decide that we want to produce a 100% neutral profile that just produce a 100% realistic result (actually many manufacturers have such presets) the looks would still differ to quite some degree, and not mainly due to hardware differences, but because the problem is without solution and requires quite some "psychovisual" creativity to get to a good approximate solution, and thus different manufacturers will come to different conclusions.
If we look at what color scientists do they stopped messing with tone curves more than ten years ago and started with spatially varying algorithms, what we call "tone mapping" in the photographic world, because subject-dependent spatially varying algorithms is required to make something work well for all images, over the whole image surface. There are various reasons this is not a good idea to incorporate in general-purpose photography though, so there we're stuck with tone curves and making the best out of it and even if we have the same goal we will end up with somewhat different results.