The degree to which people expect a RPG experience to strongly map to its more native media forms varies considerably; this is likely why you have a number of supers games that come later which don't match up as well in your view, because they were designed assuming some native genre elements were not wanted by the players. In fact, I'd argue the ongoing evolution of games in the genre is a process of iterating toward what people perceive as the market's desires here (along with some games that stake out particular levels of this and just decide that's where they're going to stand and the hell with it, but I think a game like M&M shows an attempt to push certain things closer to the native genre in some areas while consciously avoiding doing so in others).
I think the problem I have with this is that you're presupposing a level of design consciousness that hasn't been present reliably in RPGs pre-2010 (though it becomes increasingly common since the early 1990s).
Like the "not wanted" genre elements and "perceive as the markets desires". I think the vast majority of Supers RPG or Supers expansion designs over gaming history were way more naive than this. By naive, I mean, the designers weren't even thinking about those things. They weren't considering and rejecting stuff, they just weren't thinking beyond simple representational design. That's the big issue with a lot of Supers designs. They concentrate very hard on finding ways to "represent" a huge range of superpowers within a system, but they don't think "maybe the system itself is what needs to change". The system is seen as a constant in these games, and even when it does change, it's not typically due to a perceived need to make the game work better for a genre, it's usually to correct some sort of numerical exploit, or slightly streamline some particularly clunky mechanic.
The market is complicit in a sense that relatively few players were asking for more, but the idea that designers thought "Well, we could make this more like the actual genre, but no-one wants that!" falls flat to me, because there's no evidence that's actually going on with these designs.
And I think supporting my argument is that post-2005 the number of games taking the "here's a system, let's just use it for Supers regardless of whether it makes sense" declines increasingly steeply. Fewer and fewer Supers RPGs don't work pretty well at representing comics and movies. They may do so with some pretty wacky mechanics, but they tend to produce games that look like comics/movies, which a lot of earlier systems totally did not - or, as I pointed out, ironically looked like deconstructions of superhero stuff, like The Boys. Stuff like HERO/Champions is still around but it's not going to be the system of choice for anyone looking for a superhero RPG now, just something people already familiar with that like.
(I do agree that HERO did make it easy to knock people out - that seemed to be the one real concession to genre, and a conscious design element. It got in because so many supers had no-killing codes, and it's really obvious if an NPC is dead or not, so even in the naive design era, they took that into account.)
I think I'd characterize it who want that style of story's output, but not its input. The output interests them, but not the tedium of setting it up. Frankly, I'm often that way about investigations.
I don't think it's quite that simple. It's not just output, but also vibe/style. With heists it's particularly obvious, because they want the Oceans 11 or the like, but without the planning, just the actual heist sequence. I don't think that's really just "output", because it's also the process. It's more like procedural vs drama. Some people want to do most of the thing, maybe not in minute detail, but they want to do it in order, have one thing lead to the next, and so on. Others just want the dramatic bits. Neither is inherently superior/inferior but they are supported by different rules approaches.