[Christ]

I second the higher compression. Once you get so lean, you actually need compression and hotter plugs to get the mixture hot enough to ignite. Ignition temperature goes up the leaner you get, after a certain point.

Have you considered setting the engine up to make enough power normally to do what you want, but using a second program (similar to going open loop on WOT) to get power when you really need it?

Shave that head, and study your cam profile to see where you can keep dynamic compression numbers up. Consider your RPM range, and make note of the max RPM that you ever use, and keep in mind where your turbo will be most helpful. Also, keep in mind that the more you lean your mixture, the less your turbo will want to spool, and the higher compression you'll have to have to get a similar effect from the turbocharger.

Longer stroke will get you more low-end. You can use a D17 crank in your engine, but you must also use the D17's oil pump and dipstick tube, I believe.

Attempt to defeat parasitic losses in your engine by pulling vacuum on the block, to the tune of ~15 inches, IIRC. You'll want to pull it from the valve cover area, to lessen the amount of oil you get in the stream. You'll also want an oil catch can/separator.

Also, consider dyno tuning, if you have access to a dyno without too much cost, and tune your accessories to load the engine to put it in a more efficient area of the BSFC map if you're not quite able to get it there under normal operating conditions. Sometimes, adding just a little bit more load will get you "over the hill", I think, so the net effect is a gain.

If you've got the cash, re-cast your pistons so they fit the combustion chamber better. Larry says (Widmer, you know him.) that the ideal combustion pattern is a football-esque shape with the largest concentration of fuel/air stratified about the spark plug, and making sure that there is no extraneous pocket of mixture anywhere. That means fitting the piston as tightly to the head's design as you feel comfortable.

Speeding up your swirl (staggered intake events) should also help dramatically with increased detonation resistance, according to Widmer's research. If you have the means, consider staggering the intake valve opening events of the same cylinder a few degrees apart from each other, so that as the air really picks up velocity on one valve, the other is starting to open, which should induce high-speed swirl, as the second flow will influence the first one.