Your BG levels are the ultimate proof. Regardless of what I or anyone else says, if you can prove to yourself that something works, then do it. My point above is you can't rely on your testing if too many variables are involved and you can't repeat it. And there are so many variables
If I was going to test cinnamon, I'd eat a standard meal for a few days and record my post prandials every hour for 4 to 6 hours. I'd try to keep all activity the same. I'd then do the same testing with a bunch of cinnamon added. I'd do this for at least three days, maybe more. I'd probably not test for any days where my BGs weren't very close to target. Trends can be adjusted, but IMO if you are off, there's some other variable at work you haven't accounted for. My DP for example can wreak havoc on my whole morning if it decides to have one of those sunspot days -- I just don't do special testing those days.
One variable that's difficult to remove is the meter variability. It can hide any small tryue fluctiations so they are unrecognizeable. I'd almost do three readings each test for something like this. I'd then plot everything on graphs and look for obvious differences.
