I trust fully the computations relative to the ENOB/SINAD.
Regarding SNR and THD, i.e., all computations that deal with harmonics, I am less confident, because harmonics are computed on the basis of coherent sampling (i.e., harmonics are definite, one-bin spikes, which you only get by sampling an integer number of periods; the program does automatically this normalization, but it won't work very well with a noisy signal like our FMC ADC 100).

David