Yes I have a HT-P (230V version) not a B. I am using a program that tries to keep the heater at max all the time (short segments with maximal target temperature to keep the distance between target and actual HT-sensor temperature maximal). During the roast I try to reduce the heater by reducing the target temperature towards the actual HT-sensor temperature. Not that easy.
My last roast (not an Excel expert as one can see):

o temperature: 22,8C/71F
o beans: 150g espresso blend incl. lots of robusta from a local roaster to be able to compare
o weight loss: 15,3%
o Variac at 100% heater: 230V (kept constant)
o insert ET: 149C/300F
o insert BT: 132C/270F
o BT turning point: 73,6C/163F
o BT 290F/143C at 4:19 and BT 360F/182C at 6:33 (similar ramp to your Kebado roast)
o Yellow BT: 300F/149C
o fan of until FC, successively turned on 1 and 2 during the rest of the roast
o turned heater down at FC
o dump @ first signs of SC
o MET: 232C/449F
Although I ended up with the same MET as you, I obviously lost luck at the end of the roast as the ET dropped (even below the BT at the very end). Comparing this to your roast I wonder why your BT turning point is with 210F that much higher than my one. Do you have a certain trick (or some magic) to insert the beans faster than I do (maybe without opening the chute).
I did some cupping (while standing on one leg) 24h after the roast to compare the result with the roasted beans from my local roaster. It seems that my roast has more bitter flavors while the commercial one seems to be more balanced (or dull?). However, my taste buds are not very trained. Did not yet produced an espresso with this.
I found the following thread on the HT fan:
http://www.home-barista.com/home-roasting/question-about-hottop-km-8288b-fan-t9853.html#p112053It seems that Randy G. is using the fan more like I do currently that is inverse to you. However, I can not see a convincing conclusion in that thread.
How would a designed HT fan experiment should have to look like? What about 5 equal roasts keeping the fan constant during the full roast at one of the levels 0,..,4? By "equal" I mean trying to keep the same insert/eject and heat reduction points. For the pre-FC one could be able to prove that F4 maximizes convection and for the past-FC one could prove that with F0 the fine-control is easier. What do you think?
Cheers,
Marko