"How precise need intake valves be set? Exhaust valves?"

Well, it does technically affect the amount of overlap. When the exhaust is closing and the intake is opening a slight change in clearance on either valve could shrink the overlap down. But really, as long as they aren't *really, really* loose (like 1/8th inch) then it is going to be a pretty subtle thing. (However, I think you'd get a better compression test result if you slackened the valves considerably before the test to try and eliminate the overlap).

But in reality, yes, dynamically vibrating and oscillating parts when they are all rapidly moving probably introduces a certain degree of variability to the situation. And trying to adjust the valves to a degree beyond that variability doesn't really accomplish anything. Tighter valves will in theory increase the overlap up to the complete as designed amount (although the design probably accounts for a certain amount of total lash between both valves), and increase the rpm range slightly where the full effects of the overlap kick in. Looser vavles will decrease the overlap, and slightly decrease the RPM range at which the full effect kicks in, and slightly increase low rpm torque.

I'd be interested in seeing dyno tests where the only changes are the valve settings. .015, .020, .025. See if anything can be noticed on a dyno. Basically the cam lobes are pretty steep, when they start pushing it open .015 vs. .025 gap really won't make that much of a real difference in how far open the valve is when the cam is at a certain degree. Technically speaking, yes, it changes the actual overlap point. But in reality - I doubt if it really matters that much if a valve is open 0.0 or 0.01 (the difference between .015 and .025). And a couple of degrees later I doubt there is any real difference between a valve that is open .115 and .125.
--
I'm JohnMc, and I approved this message.