I think another factor that enters this discussion is the amount of time that the ignition induced pressure increase has to apply its force to the piston. At lower engine speeds (say 1000 rpm) the time is relatively long when compared to time at higher engine speeds (say 3000 rpm) and even less at maximum engine rpms. As a result there are diminishing returns to the earlier initiation of the pressure increase and as noted 1 degree to early has significantly more influence than 1 degree to late. The pressure build after spark starts the pressure build takes time, milliseconds but still time, so at some engine speed that earlier spark timing doesn't generate any benefit.
