I think a 5% return is approximately optimal in the early stages of most games - that's about what you can get from building settlers in monarchy. GOTM 68 is rather artificial (one settler + all the techs) but it has shed some light on growth rates under Democracy.
Magic Gorter managed to double every 10-11 turns [for about 36 turns] which is a growth rate of about 7%. He used normal ICS methods, and 100% taxes to pay for rushbuys. I suppose he could have sustained this growth rate quite a while, especially if given a few happiness wonders he'd normally have with a democracy.
He did not indulge in trade, but in a normal game, that'd probably give even more than a 7% return. Not sure... I haven't tried to estimate the typical costs, benefits, speed of trade. In one PBEM we figured it was as much as 15% (but movement was faster in that scenario, which inflates the return).
In GOTM 68, I grew slower at first [also in Demo], but after celebration-growth kicked in, I got almost 7% [while also conquering the AIs]. It's not clear whether to measure growth in terms of number of cities, Civ2 score, or Mfg, etc.... but I think it's clear that you can get even better returns from building engineers in Democracy than settlers in Monarchy.
This makes investments in factories, King Richard's, etc even less attractive than before. Investments in irrigation and mining are probably better than before (because of engineers' 2x speed) but without careful analysis, I'd guess they usually don't pay off very well (for most purposes, such as EC, EL or GOTM scores).