I agree, but here's a quote from the CEO explaining (to some degree) the choice:
“So we did actually ask AMD about this the first time they told us about Strix Halo. It was literally our first question: ‘How do we get modular memory? We are Framework after all.’ And they didn’t say no actually, they did assign one of their technical architects to really really go deep on this—they ran simulations, they ran studies, and they just determined it’s not possible with Strix Halo to do LPCAMM. The signal integrity doesn’t work out because of how that memory’s fanning out over the 256-bit bus.”
I never understand this justification. If the tech does not allow, then don't do it if it's against your own core value. Instead, they join the quick cash grab AI-train.