Cool stuff! A couple thoughts, if you want 'em - depending on exactly how you want the engine to make power, a stuffed case may not be needed, and may be detrimental for power band (I think the guys that developed the Aprilia RSA125 ended up running pretty large crankcase volume because it made wider power?). And compression ratio/compression readings don't matter a ton when you're going for top power, because the pipe can stuff plenty of air in for it not to matter a ton.
My 50cc Peugeot has a pretty low compression ratio and large crankcase volume, and makes zero power below 8000rpm, but it makes a ton of power from 9k to 13k (13hp or so if I had to guess? Does over 65mph and accelerates quite well, pulls 5% grades at over 50mph. Gotta get it on a Dyno someday). I have the luxury of a well-tuned clutch and variator, so I can get away with a narrow powerband with low compression and hard-hitting pipe. That low compression is a bit easier on the motor too, I run very very low head temps even when ripping at top speed and climbing long hills. I almost don't trust my gauge because even after climbing up a 500' hill at 10% grade doing 30+mph, it says I usually hit about 140F. But I can easily keep my hand on the radiator, so who knows?
Things obviously change for single-speed applications, but you need tons of power for drag racing regardless, and with a centrifugal clutch, your launch phase (while clutch is slipping) is determined by the torque you make at the clutch slip speed. F=m*a, P=F*v, so at any particular speed, more power (to the wheel) gives more acceleration - which sounds obvious, but it means if you can downgear to let the engine rev higher and get into the powerband (especially after the clutch slip phase), you will often have more power available to you (even if there's less torque!). For instance, if you shift the powerband up a good amount to make more power, and sacrifice some torque to do so (let's say 10% less torque, but 20% more power), you can change the gearing by maybe 20% as a result. So the net impact is better performance all around, even when the clutch is slipping (roughly 10% better acceleration even with less torque).
My real 2 cents, it you're taking this much time and effort with this, it would be well worth it to get some good simulation software like EngMod2T. Costs less than your bearings, and has the chance to make a much bigger impact! 50% improvement on 99% efficient bearings only gets you to 99.5% efficiency, so a 0.5% overall improvement (even if you have a noticeable improvement in revvability). You can easily find several percent with better porting and pipe than what you get off the shelf.
Anyway, I'm looking forward to seeing more of this project! Good luck!