I plan to do a test comparison of oil pump settings on a Polaris 800 EFI, which should be interesting. Dyno testing various fuel/oil ratios on carbed engines is difficult because mixed oil displacing fuel through the main jet can result in leaner (more powerful) mixtures, just like installing smaller main jets. Plus the more oil mixed with gasoline, the higher the specific gravity, too, further skewing results. Lots of people have tried various % fuel/ oil mixtures here on carbed engines without much discernible difference. So doing this sort of evaluation with fixed EFI fuel flow and varying oil flow with injector pump will make this practical. My gut feel is that more oil than absolutely necessary for bearing reliability will reduce HP due to excessive friction in the crank/ rod ball/roller bearings. Kevin Cameron reminded me that it might be like driving down a wet road in your car, and when you encounter a sheet of standing water the car suddenly decelerates from the added friction.
And do we really need more oil to create ring seal? My opinion is no, especially since creating nearly perfect ring seal on worn engines has shown us disappointing results. Remember when SkiDoo had trouble with crumbling moly in the rings on the 600 twins? For a DTR test (and to learn something) we dyno tested one with bad rings (25% leakdown as I recall), then relaced them with new rings (less the 5% leakdown) and got almost zero power improvement. In addition, it seems that the NASCAR and F1 engine people strive to remove all but a few microns of oil from the cylinder walls with perfect oil control rings to keep the top rings just barely lubed, but not dragging down HP.
This discussion makes me think back to the comprehensive dyno evaluation of an aftermarket two-stroke oil I did with my XLT600 (posted on my website). As recommended I had soaked the cylinders overnight in this stuff, then put several hours on the engine with the new oil, only to see a 2% loss in horsepower on an engine that repeated with a few tenths of a percent. Fortunately when I drained the oil tank and refilled it with Polaris oil, the power immediately came back. It really didn't make much sense then, but could the aftermarket oil have had higher viscosity, hence higher friction? On four-stroke bikes and sleds, we find as much as 4% horsepower increase when we use 0W20 oil instead of thicker stuff like 5W30 or 40. Maybe thinner (and diluted more with gasoline) injector oil will help us make more power. So when we add 10HP to a CFI4 800 by leaning out top end fuel flow by 10-12% we are making the oil/fuel mix that reaches the top end even higher viscosity? Can we make even more power by dropping the oil consumption by 10%?
But then we seem to be searching for better longevity of parts as well as added HP. Can we have both? If it turns out that cranking up the oil flow adds power, then probably yes. And those who have suffered from parts wear, adding oil seems to be smart because who cares if you lose 2 HP if you can make it through a season on the same pistons. (However, on some sleds that exhibit premature piston wear, just proper PCV midrange fuel tuning can eliminate that, which is way better than dumping in extra oil). But what if we make 1% more power with @20% less oil? The greediest of us will surely crank that oil flow down if that gives us 1.5hp more than our pals, but will we have more wear on other parts?
I should be able to find the optimal pump setting for max power, but it may be minimal. But it does seem that having the most HP means having the least friction which should = less wear on parts. More questions that need answering! It will take the dyno to determine max HP, plus 1000s of miles of field testing to determine what is really optimal. I hope be able to provide the first half of the equation-- then let the greedy ones be the guinea pigs!