Super interesting Luc.
First thing to note, if a default gets you close when you drop in an ORTF array? - if it's too different to your space/piano, what things do you think may be things you'd do 'at the desk' maybe.. compression/etc..
Not sure what to offer tho - but if you have MP3 audio of the type of results you get with similar piano in a similar space, maybe that could establish a beam on some things you could try. (you may know the upcoming space or piano so maybe some audio could match near enough?)
Probably many tiny things (apart from all the usual obvious things, favoring 1st overtone a touch, or opposite I find is an overlooked trick for helping 'various mics' sound more or less realistic depending on reverb thickness, distance of mics - lots of moving parts even for a simple preset).
One of my likings for ORTF is being able to get a clear 'almost close' sound from safe distance, picking up a nice room ambience - but you may love further back or closer (haven't done this outside of Pianoteq for many years tho)..
Adjusting things against an audio reference would be the best for sure of course - but since the performance is 'on the night live' so to speak, it's a doozy of a challenge to get that right
(or do you want to adjust the Pianoteq to improve it after performance perhaps too? as in, provide something as good as possible for the event, but improve it against referenced audio afterwards? Would that be cheating for this experiment?).
So many questions - keep posting Luc - hope to help if I can.
Definitely would love others making FXPs to join in too - this is a pretty cool learning curve situation for anyone interested in FXPs.
Pianoteq Studio Bundle (Pro plus all instruments) - Kawai MP11 digital piano - Yamaha HS8 monitors
