A major problem with that is that the appearance of servant economies being more productive than extractive economies has more to do with the fact that extractive economies get the short end of the political stick. When American software developers get paid a hundred times as much as Chinese assembly line workers, who in turn get paid ten times as much as Congolese miners, it is far more because of the uneven distribution of aircraft carriers than because of the uneven distribution of productivity.
So while a planet full of financial centers within a well diversified empire should be richer pro capita than a mining outpost, an empire which has nothing but financial centers should not have a ludicrously high GDP. It should be in complete economic paralysis. But Aurora doesn't model that, because its political, international trade and intra-empire cash flow models are fairly rudimentary.
This discussion also touches upon another fault in the handling of taxation, government budgets and generic "wealth" mechanics: Governments don't get to "save up" money for a rainy day, the way Aurora lets you do. Most of what the government wants to buy with its money is labor, and labor dissipates immediately if it is not used.
Not-spending money today therefore does not allow the government to spend that money tomorrow without straining the economy. In the real world there are no little gray men in suits smoking rose petal cigars and running a time savings bank where the unemployed can stash the man-hours that the government not-buys. Conversely, spending money today does not require the government to not-spend tomorrow to avoid straining the economy. (There might be time-lagged higher-order effects, but they all show up in the output gap, so given sufficiently alert and fast-acting fiscal authorities you can safely ignore them in your macroeconomic planning. )
A more elegant solution would be to have a series of "baseline labor force participation" techs, and simply let empires start at different levels of those techs and different levels of the GDP pro capita techs. This might also open up possibilities of modifying labor force participation rates by governor bonus and by previous employment history (persistent labor shortage brings people out of retirement, provides better career opportunities for academic drop-outs and therefore encourage earlier labor force entry; conversely, persistent unemployment does the opposite, and also pushes people into the informal economy, from which they will take some time emerging when unemployment drops).