Use external perimeter acceleration for test acceleration

[npburns224]

I see that the pattern generator is choosing the max of perimeter_acceleration, infill_acceleration, solid_infill_acceleration, top_solid_infill_acceleration, and external_perimeter_acceleration for the test acceleration value.

I'm using a Prusa MK4 with a 0.4mm nozzle. When using the default 0.2mm STRUCTURAL print profile, the external_perimeter_acceleration is 1500mm/s² and the infill_acceleration is 4000mm/s². The difference between these two values is fairly large.

Wouldn't it be more ideal for the test to use the value for external_perimeter_acceleration since those are the most visible part of a print? Or perhaps an average of some subset of the acceleration values?

I may have some misunderstanding regarding this particular tuning process, but up to this point I've always used my external_perimeter_acceleration value for generating test patterns with the Marlin LA calibration tool or Ellis' PA/LA tool.