Lots of people have been using VNT for ages. It's nothing new. Borg-Warner just decided to use expensive materials for them, apparently. The previous problems had to do with vanes which could not withstand high EGT. Materials have advanced since then.
12:1 compression and 17 psi of boost means little. Not only is static compression misleading, but "boost pressure" is even more misleading.
If you want to compare, check air-mass/combustion and BMEP. Given the 260 lb-ft of torque from a 2.3L, I gurantee that BMEP is quite low. Saab have been running more than 260 lb-ft from 2.3L for quite a while. The static compression of those Saab engines may only be 9.3-9.5:1 ... but they are not variable valve timing engines.
What Honda has done is alter the valve timing so that, when the engine is under boost, the engine "acts" like it has about 9:1 compression instead of 12:1. (Like someone mentioned about dynamic compression being the one that matters.) But when the vtec is in "cruise" mode, it "acts" like it has the full 12:1 comression ... thus allowing good gas mileage.
Pretty simple really. I figured they'd do this a few years ago. Nice to see they are putting it into use.