In real basic terms it the number of degrees that the intake valve is open passed top dead center, but other things like the profile of the cam, and the LSA, which has to do with the overlap between the intake and exhaust valves comes into play too. It easiest to use a calculator like skyhook was talking about. Thats why you want to make sure you components work together, so you get the corrected CR right. Here's a comment I received a while back from Mike at Latus in Portland, doesn't really explain how to calculate it, but talks about the affects.
You are correct in that an earlier intake closing will make more cylinder pressure vs. a later closing with the same mechanical (static) compression. The two magic numbers we shoot for are corrected compression (based on intake closing and static) and cranking pressure (what you would see on a test gauge in the spark plug hole while cranking the motor). Our experience has shown (at least here in Portland, Or.) that corrected pressures above 9.5 when using 91 octane pump fuel cause detonation (pinging) problems in hot summer climes or when traveling. This is something of a personnel tuners/builders preference and others may choose higher or lower numbers base on their experience and area. The cranking pressure seems to be all over the map and is a HUGE function of how it is done and type of gauge used. I use a Snap-on Motorcycle test gauge and based on the numbers it gives, I like to see 195-210 cranking pressure (warm motor [not hot], throttle wide open, good battery/starter speed, spark plug out of opposite cylinder). This gives best torque without pinging for street applications running on pump fuel. Numbers below 170 cranking will produce a lazy torque curve due to lack of cylinder pressure. Lighter bikes can stand slightly higher number and heavily loaded bikes may want slightly less. Actually two measurements of the same thing, one direct (cranking) and the other calculated (corrected). Because of the variables in the cranking test, we prefer to use the calculated method and then confirm with the cranking.
Scott, you did a great job of documenting your build, I may have some questions for you soon. I plan on doing my RK this winter.