Well once you get your EGT's on there and learn exactly what the reports mean you can eliminate at least your fuel to air from the question. What year head do you have on it? Seriously cant imagine how it isn't detonating wildly with pump 91 unless you are riding at 9,000 or up on a stock 09 head. There are extremely varying performance gains with the correct head both compression and cut style.
What does CPI say their pipes run best at RPM wise? On ours 8450-8500 is magical and there is such a huge difference between that and 8000 that you can hardly believe it! A little concerning that your valve springs would earn so much top end RPM. For me all they did was change at what RPM they opened at. It's my understanding that they should be fully open at that point not partially. Might be identifying a problem in the pipe design right there? Also note for some strange reason these motors systems with (p85 to team roller) don't seem to show an over spin when you have way too light of a primary weight. I wonder if that is why you are getting the same top end R's on 4 grams weight difference and why it is having that stronger pull 3/4 to almost full then feeling done on the very top end. Mine acted like that with the 2010 head and the pipe sensor coding so that it went to 10% fat..then we fixed the sensor and ran back in a good state of tune but the 2010 head performed poorly in comparison with the 09 head. It had that exact same feeling of no gain pegged over almost pegged as well as some settling. When we went back to our original 09 head it goes right to 8500 and stays there as long as you have the stones to hold on. Anyway all of that is theory and speculation until you have your EGT's on and have exact knowledge of your tune at that all out range. 100% I can tell you just that 10% pipe issue costs 400 RPM's minimum. Not surprisingly when I messed with hooking and unhooking Eth wires we saw about 150 to 200 RPM's and a 100* pipe heat. IE half of the pipe sensor enrichening so that makes sense since the eth wires are supposed to enrichen 4-5% when disengaged. I don't know how extensive you got into plug reading but it is very very finite and unless you have alot of experience there isn't much plug reading variation between those reports. I went at it just the opposite in that we identified the ideal heat we wanted by plug reading and EXTENSIVE experience on the system by people way smarter then me, then compared those known reads to our instruments reports. Maybe (hopefully) youre still just a little fat yet on the top end. Mainly since we'd assume a manufacturer would test their product and have some success with it in your given arena so that they can advise you!
Now also note on your clutching there is no way your sled will pull 8400-8500stock Polaris 10-66 at 8000ft with 91 fuel on a stock head. See if you can beg borrow or steal a little more helix angle then stock...try for 64-40-.46 and the stock 800 driven spring. If you are willing to put a higher compression head on then Pol 140-330 primary with 62's and that helix should be right there for you at 7,000 to 9,000. Or maybe a 140-320 TEAM RED if you are going to keep the stock head and you may need just 60's. Depends on your track, weight, and the conditions of the day.