Lower chamber pressure doesn't result in lower Isp for a vacuum engine. Instead it is just is proportional to the gas exit velocity, which is related to the gas temperature and molecular weight. The reason why spacecraft use a relatively high chamber pressure even in a vacuum is to get the same thrust from a smaller throat area, which allows the nozzle and everything else to get smaller, which improves the mass fraction and packaging convenience even if the Isp difference is negligible.
For a rocket in atmosphere, having a lower chamber pressure does hurt the Isp.
https://space.stackexchange.com/que...-impact-on-isp-for-different-types-of-engines
You should think of propellant mass in stages above the currently-burning stage is just like any other dry mass in the currently-burning stage. If you can reduce the propellant mass of the third stage by improving its Isp while keeping everything else equal, it acts just like reducing the casing mass of the 2nd stage during the 1st and second stage burns, in addition to improving delta-V you get out of the third stage. This compounding benefit of efficiency in the upper stages is why so many orbital rockets go to the trouble of using a LOX/H2 upper stage like the high-Isp (~450 seconds) Centaur upper stage, when most of the impulse is provided by a high-thrust, lower-Isp (~340 seconds), LOX-kerosene-burning first stage like the Atlas and Delta series of rockets.
The ideal mass ratio between stages depends on implementation choices like how you trade aero drag loss vs. gravity loss, and how efficiently the different stages package together.