Looking over this, I think that you're outlining a point somewhat (not completely, but somewhat) orthogonal to the issue around hit points trying to model two different things at once.
Now, hit points are very much gamist in function (though I wouldn't say simulationist; while I won't speak to how Ron Edwards used the term, my conception of it has always been about mechanics that set/affect the narrative directly rather than mechanics that happen to abet dramatization), which is why I previously pointed out that they were an area where simulationism backed off. That was, as I posited, a consensus (or maybe I should have called it a compromise) that everyone was unhappy about, but could live with.
But while the issue of scaling wasn't necessarily limited to healing spells being less effective as a character leveled up (that was just the inverse of the aforementioned issue of how the same 8 hit points' worth of damage in one attack could kill a commoner outright, but meant little to a high-level character), 4E's attempt to fix that by having a central healing mechanic that operated on a percentage basis was a legitimately good idea...one that it completely undercut by leaning hard into having hit points (or rather, the loss of hit points) be a model of being progressively injured until your life was in danger and simultaneously being a model of progressively losing combat capability.
While there was a modest amount of conceptual overlap in those two metrics, they were still dissimilar enough that they caused a cognitive gap for a lot of players in how a given solution (i.e. a warlord yelling at someone to let them use a healing surge, a cure light wounds spell, etc.) functioned for both – since, again, it was a single mechanic modeling two different things at the same time – despite being presented as a fix for only one of those two things.
The result was that the percentage-based healing solution was presented as part of a much greater problem.