I'm not sure. We may not have.
But TKD glosses this construction as "A's Q is many, B's Q is few" with a possessive construction.
It may be notable that it's not e.g. jaq la' law jaq yaS puS for "the commander is bolder than the officer", even though "the commander is bold" would be jaq la' and not la' jaq.
Perhaps the construction is similar to things such as jIH 'emDaq "behind me", which is "at my area-behind" with a similar "possessive-like" construction that does not use 'emwIjDaq (in the standard dialect).
But you may be right that we have no canon examples and that this is (speculative) extrapolation.
Maybe something to ask Marc Okrand.
There's another in paq'batlh:
luqara’ ‘oH pongwIj’e’ jupwI’
bIralqu’ ‘e’ lu’ang mInDu’lIj
SoH rallaw’ law’ Hoch rallaw’ puS
My name is Lady Lukara, my friend,
Your eyes show more bloodthirstiness
Than I have ever seen!
This one also demonstrates how to indicate doubt in a superlative construction.
I'm sure André is referring specifically to having a first or second person be the A or B in the formula. Most of the examples we have use a third person in the A and B slots and thus, if the verb did take prefixes it would take a null prefix. Thus, how would we be able to tell if it were "inflected" or not? I agree with you and feel that the common examples are not using a null prefix, but rather that this formula is not inflected as you say. Do we have any canon examples of first or second person comparisons to support that?