Dyno'd my 2.0S today.. What the h#!!

[Staylo]

Assuming you have no other fuel restriction issues, I'd try the 110 air correctors.

[rsscotty]

I have seen this lean condition at higher rpm range before on the dyno. It turned out that one of the booster venturis was installed wrong. They will press down into the carb. main body either way you can install them, but the internal tube or horizontal wing inside the booster venturi has to face the direction of the air corrector jet. See photos below. You can look down each throat with a flashlight and mirror and you should get the idea.

[ValveFloat]

I think it is best to separate your question into two parts:
1. Why am I off the scale lean above 3700rpm?
2. What can I do about the mid range flat spot?

1. Where was the AFR measured? If it was measured at the tail pipe with a probe, it is possible the probe came out of the pipe enough to be sucking some fresh air. I know from experience that it is tough to get a probe far enough into a stock 911 muffler to avoid that happening. Like someone else mentioned above, your car would barely run at above 18:1 air-fuel ratio.
My advice would be to completely go through the carbs, make sure everything is correct and clean, check your fuel pump for sufficient flow, etc. Have O2 sensor bungs welded into your headers/heat exchangers just after the collector and return to the dyno. The stock jetting should not be that far off.

2. I struggled with a mid range flat spot very much like you have for years with my '68 911L. I tried all kinds of jet/air corrector/exhaust changes to little avail. As mentioned above, running larger idle jets than would seem necessary did help. Maybe this is due to the differences in fuel today vs. the '60s? Eventually I looked to the ignition system and that is where where I found the best results. I suggest getting the distributor advance curve measured to make sure its correct. Make sure the ignition system is healthy overall.
What exhaust are you running?