If true, then the NSA isn't that influential (Score 1)

Conspicuously missing in his blogpost is SP 800-227 was recently published by NIST September 18th, allowing for hybrid ML-KEM/ECDH key encapsulation to be used, even going so far as to allow the ECDH component to use curves that would not be FIPS approved if they were standalone, using x25519 as an example. I'm as skeptical as anyone of these agencies given their track record, but if we assume this narrative is true, then it appears that the NSA failed in the blocking of hybrids as FIPS approved, and hybrid algorithms (even using non-NIST curves!) are considered to be approved in FIPS validated cryptographic modules. Admittedly, I'm less knowledgable of the politics of IETF, but with respect to NIST it appears the NSA hasn't even won this supposed battle on their home turf.