I would get a BT tool if you have not done so. Are you sure you cannot log AFR? No Lambda probe value before the cats on your logging tool?
Anyway I have no idea how 10 degrees of timing retard translates to power but losing 10 degrees throughout the entire powerband I would have to think equates to a significant amount of power lost from potential peak power.
I don't think it is surprising however as from my research these cars so rarely if ever hit their timing target in real world driving, especially on 91 and often not even with 93.
These are the kind of objective data that really illustrates that while on a controlled dyno with perfect conditions and a couple pulls 91 may hit peak power, in real world pulls, 91 simply does not come close to cutting it in terms of making the stated power of this engine.
I still get knock using 94 octane but like I said a tuned file is a bit different but hitting upwards of 34 degrees, in the higher RPMS I still get some knock above 8k with 94 so you can imagine straight 91 you would be nowhere near able to hit even stock 32 targets
I wish more dyno places would include some basic logs with a dyno. A dyno plot with some timing info, IAT and AFR (which some do) would really allow more accurate comparisons across different dynos.
Would explain why one car with the same mods may be putting out more/less than another identical modded car if the timing on 1 of them was a lot different for example.