Civilization-3 handled budget sliders on an empire-wide scale, with a simple percentage of income going to research, another going to morale, and any leftover percentage that you didn't assign going to the treasury. You could also go into individual cities (planets in our case) and assign people to different jobs--tax collector raising money, scientist increasing research, entertainer boosting morale. The combination works quite nicely.
A counterpoint about moving all finances around: We are moving only those finances that we can move, and the sliders are there so players can decide how to spend their money. There is a sense of worry over moving the sliders "too much" and I think that is a healthy thing. It should be possible to do exactly that. For instance, if I decide as Grand Poobah of my race to stop all expenditures of money on ship manufacturing and devote my entire economy to research, then there will be consequences. The enemy will be pumping out ships while I race up the tech tree.
One place that I've felt comfortable with modelling wasted effort is rush-production queues, where you have the option to hurry a given item/queue along for a price. I liked in SEIV where he did rush-job production queues that increased the cost, and then charged recovery time penalty of building at 1/4 production rate for an equal length of time. Then there was the hopelessly convoluted MOO3 which tried an economy-overdrive slider on their worlds and for me it was just a chore tweaking it to the edge of a color change, while the other sliders changed in oddly disproportionate ways, then erased my decisions a few turns later. I wasn't a player in that game, I was just one more hint for the simulation. I give em credit for guts though, because they didn't just pump out another formula game.