So, in effect, you are saying they voice them so lifeless and flat so that they sound good out of the built-in speakers (which would have to be a LOT more powerful and better frequency response to sound as good as a full system)..?

Yikes!
Why not make them full and punchy to START with, and then use the comps and EQ's to flatten the sound out for the built-ins, rather than the other way round, which is MUCH harder to achieve?