Right, but was the dyno lambda sensor in the tail pipe? If so, that's why it was showing lean.
I'm almost positive that VTEC engaging too early isn't what's killing your torque. The curve looks fine at and past 4000rpm. If VTEC was killing the power, one would see a sudden dip at that point, but one doesn't. If anything, I would guess that VTEC was NOT engaged on that dyno you posted. At 5500 rpm (stock VTEC engagement point) the torque starts to drop off. Considering the stock torque peak is at, what, 6500rpm? it shouldn't do that.
If it's not that...your adjustable cam gears are installed, right? If so, maybe they aren't zeroed and are dialing out some valve overlap to begin with?
The only other option I see is that there is RIDICULOUS heatsoak that accumulates through the dyno pull, which kills the top end. However, I'm leaning toward VTEC not working. Try setting it to 2000rpm, or something really low, and see if it makes the engine run like a dog out of boost. If not, then something is wrong.
::EDIT:: Actually, I was just looking at some other JRSC dynos and I think that the torque dropping off is maybe just a function of the supercharger spinning so fast that it's out of it's efficiency range. Nothing some spendy porting won't fix