I'm not sure I'm following. I guess it depends on the relative subject brightness of the whites of the eyes and hot spots on the nose/forehead. If they are very close, they will end up on the same part of the film curve. If they are further apart I suppose you could expose/develop to push only the brightest highlights onto the shoulder.

To me this still sounds more like it falls under print controls (ie selectively darkening certain highlights relative to other highlights even though they may have similar densities in the negative).

But I guess what's confusing me a little is when you're talking about using the shoulder but also saying you want to retain more detail at the very top. That's why it ends up sounding more like a printing control issue than a negative issue.

Regarding the specific films, the three you mention are nearly the same until you get to the extreme highlight densities. TMX 100 for example, tends to shoulder a little more gradually (ie it starts a little earlier) than Delta 100.