I think there is also some awareness amongst the engineers of old, less so now, about the devices and locations and manners in which music will be played.
The aforementioned DSotM spawned a couple of singles and not only does that reflect what, musically, might work when separated from the album (singles off concept albums is a complicated business), I think it’s reflected in the mixing.
1970s radio play would have favoured certain bands of frequencies in the mix over others, and I sometimes feel like there was a nod to that in the mixing punching up certain parts of the mix specifically for that reason.
You look at modern music, though, and there’s no need to differentiate - radio is far diminished to what it once was, and very often now people are consuming their music from devices with tinny little speakers (or tiny little earbuds with tinny little speakers) that have limited bass driving and so the engineers have shrugged and optimised for maximum noise. It’s not helped by the compression algorithms that get used (mpe and friends) which strangle out the botton and top ends of the frequencies either, again pushing the engineers to max out the noise envelope to keep that stuff in (to a point)
There’s a reason that in the headphone world I have a big ol’ pair of Sennheisers, and not just because over the ear is more comfortable or blocks out the rest of the world (though these are definitely factors), but because they have the much bigger speakers in them that can actually do something useful in terms of pushing both ends of the audio spectrum.