I'm sure this is way over the heads of most reloaders and irrelevant to those who simply want safe loads for their guns. For that, please just look up the loads in manuals.

If that's the case, please ignore this post. It is meant only for those who want to get a conceptual understanding (not exact numbers) of the relationship between powder charge, pressure and case volume.

Here are some sample loads I ran on Quickload a while ago when I was also thinking about what the OP was talking about .
Typical disclaimer : THIS IS ONLY CALCULATED DATA AND NOT MEANT TO BE ACCURATE LOAD DATA. DON'T EXPECT THIS DATA TO BE SAFE FOR YOU.

That being said, Quickload has been very reliable for me when predicting safe middle of the road loads especially when comparing one powder to another and one bullet weight to another. It does diverge when predicting muzzle velocities in very short barrels (2"). It tends to slightly overestimate 45ACP velocities (compared to what I've chrony'ed) and it underestimates handgun load velocities when you start loading hot (for instance 38 sp with a load of 5g Unique under a 158g bullet). I never use a load greater than 75% predicted SAAMI max.

Here is a plot of MV, Pressure and Energy when a 45ACP 230g LRN is loaded with Unique:

The yellow markers are what I consider the max safe loads.

You can see that in this small section, velocity does increase almost linearly with powder charge, but pressure (your safety margin) increases exponentially. That means when reaching for more velocity, you reach the pressure danger level a lot quicker as you increase the powder charge.

The next graph is a plot using the same bullet, same charge, but seating the bullet deeper each time. I think it was with a 5.6g charge:


You can see that MV does go up with increasing seat depth, but pressure goes up much faster. That's why putting 4.5g Unique under a 125g bullet in a 9mm case comes out faster with a lot more pressure than the exact same bullet with the same charge in a cavernous 38 special case.

None of these plots should be taken as absolute, but you can see the relative relationship of charge variation vs case volume variation on pressure and MV.
Hope this satisfies a little of your curiosity.