In principle, one could use metallography to distinguish between (A) gas-atomized, spherical, metallic feedstock particles not remelted in thermal spraying and (B) thermal sprayed particles of the same powder lot that melted and resolidified in flight.
The as-etched microstructure of a particle can be greatly influenced by its solidification rate. So (A) and (B) particles of the same true size might be distinguishable if the solidification rates of (A) in manufacturing and (B) in spraying differed.
Note, though, that a metallographic cross-section you see may not be through the "equator" of the particle, which makes comparisons tricky---big particles can look smaller in cross-section than they really are, while solidification rates vary with
true size.
Also note that, if the thermal spray standoff distance is too low, some particles might have too little time to melt. If the standoff is too high, some molten droplets might solidify before impact.
About plasma fluctuations: As you indicated, these can affect the heating rate of sprayed particles. They're not all exposed to the same melting conditions. Some recent equipment designs give exceptionally stable plasma jets. Here's one new example:
https://www.arzell.com/plasma-spray-system/------------------------------
Robert Miller
Materials Engineering Consultant
R. A. Miller Materials Engineering
Indianapolis IN
(317) 698-5479
------------------------------