I can get behind the idea of spies costing less the more you have (even so far as to have them decrease exponentially every million or so total spies), in addition to Spy_power=spies*acres^(-1/2).

The suggestion for making it a more dramatic -1/3 is due to the massive expense of holding high spal at high acres, which decreasing the expense curtails. It should be easier to run 200 spal at 50k acres than 5k which currently it is not. So if it were a combination of the two, the end result is roughly the same, although it does more to reward high acre high spal countries. Imo win/win. I can way get behind that.

I think decreasing CD to a constant 3% is also reasonable but not as effective as having a faster spy dr on CD. You'll still be breaking a 10m troop country at under 4m with 20 CD and 7 demo. If spy dr hits at 10 - 4% CDs, you'd be looking at more like 4.5m than a 3.8m break on a 10m troop country. Since 7% of troops is basically peanuts in most cases, i think both solutions are good.

I think you are 100% spot on Gerdler.