Using Knock sensor engine retard values excursions to validate timing changes

Since I dont have access to a dyno I have been using ignition timing correction knock ( front and rear ) to reduce knock excursions from a maximum of -6.25 degree excursions to a maximum of -2.25 degree excursions in the general range between 2000-4000 rpm between 65-95% load. Tuning an Indian 111 motorcycle

This has resulted in removing around a maximum of -4.5 degrees in various cells in that rpm and load range ( after extensive logging and looking at which load/rpm range the knock sensor was removing timing.

My question - is this a valid way of fine tuning the timing for engine protection at high load and high rpms values. I know I have no real idea of how far away I am from MBT ( I am using a stock manufactures primary ignition table as base timing). I also know that knock sensors tend to remove more timing than necessary -.and if you removed all timing as indicated by knock sensoryou would indeed be removing way too much and be way off MBT and hence leaving power on the table.

It's not a bad way of addressing knock retard however you do need to be aware that often the knock control strategy will be a little over aggressive in removing timing in response to knock. You also don't want to retard your timing map to the point there is never any knock retard showing as this will almost certainly be too conservative and cost you power. What I try and do is ensure I don't have repeated knock that occurs at the same combination of load and rpm. I'm totally ok with an occasional knock event that isn't repeatable however as that's what the knock control strategy is there to correct.