I listened to the RTi8 many times. Unlike the RTi10, they don't need much power to sound good. Anything like a HKAVR240, Pioneer 1015, Yamaha RX-V657 receiver should suffice. If you email Polk, I bet you they will tell you something similar. I also read about people measured the impedance of some RTi speakers to be more like 4 ohms instead of the specified 8ohm nominal, but I thought it was for the RTi10, not the 8. I also have my doubt about how it was measured.
If you use an ordinary ohmmeter to measured the impedance across the terminals you are in fact measuring the speaker's d.c. resistance because the ohmeter's output is d.c. That may be the reason why the measured value is lower. The impedance of a typical speaker is complex, i.e. not totally resistive, and will vary quite a bit with frequency. You can't use an ohmmeter to measure impedance. Impedance is usually made up of resistance, inductive and capacitive reactance. Mathematically it is a complex number and the symbol is Z, not R.
You question about the effect of 4 ohms: To deliver the same power to a 4 ohm speaker, the amp will have to push square root of 2 or 1.414 times of current that it would need to push, into an 8 ohm speaker. It does not mean 4 ohm speakers need more power (all else being equal), but it does mean 4 ohm speakers need relatively more current for the same power. For simpicity, consider a pure resistive load, the power forumla is:
Power=CurrentXCurrentXResistance, so if you half the resistance, for the same Power, Current must increase by a factor of square root 2. If you double the current, the power will double.
Bottom line, more power is definitely better especailly if you have speakers with nominal impedance much lower than 8ohms. On the other hand, as others mentioned, you may get more for your money if you invest on speakers, room acoustics etc., instead of on a $2,000 amp. For $2,000 you can probably get a nice pair of Paradigm Studio S2.