Hi Metuchem,
Your concern is apt on multiple levels.
I said above, " There are column types that do not lend themselves to higher plate counts, that is for certain.
(The peak in question is an early-eluter in the chromatogram--unlucky in this case. I'm assuming that there is little/no extra dead volume in the LC system/LC column. If the retention factor of this peak is 2, the minimum is achieved.)
For setting a (Plate Count N) value, I'd inject this substance at the appropriate level multiple times on different days (and different columns) to make the best evaluation."
Short answer to your last question is "Yes", you will have to validate the value of N for this peak. To do this is not difficult, but you'll require a new protocol, another CN column (at least) and multiple weighments/analyses on different days (at least), preferably with different instruments. The US FDA guideline for N > 2000 from 1994 you refer to (from Nov 1994):
https://www.fda.gov/downloads/Drugs/Gui ... 134409.pdfon pg. 28 is a Recommendation, not a mandate. If you perform the experiments, show the statistical workup and justify the value you set for the final method, you should be just fine.