I'm not a DSO photographer, but -
It seems to me that the brighter objects and their surroundings had to be dialled back
significantly in the more recent APOD so they don't appear completely blown out, and it was during that step that things went astray, for example by arbitrary selection of which areas/features to apply this to, and by what amount. I've seen this in some aurora photos, where parts of the aurora ended up being darker than the surrounding sky.
Secondly, we don't know that what's being shown in red in either image (but particularly the 1st one) is
actually Hα nebulosity as opposed to a galaxy halo, for example, where the disparity is especially pronounced. Maybe a bigger effort was made in the 2nd photo to isolate actual narrowband Hα from other stuff. There are
full-colour images online that
show the same feature but
much less red. So I'd say no, the H-alpha dataset, particularly in the earlier image, is
not accurate.
Thirdly,
this article suggests that while hydrogen is possibly abundant in halos around galaxies, it's not necessarily emitting at the H-alpha line, so mapping the whole thing to the same colour (assuming its H content can even be detected in the first place) as H-alpha would be taking significant liberties.