Every hypothesis the thesis rests on, carded in the same shape so none of them hides behind prose.
An assumption without a falsifier is a preference. There are none of those on this page.
The review begins with the smallest version of the company and asks what has to be true for the larger version to follow: that the core can stop changing, that materially different sources reuse it, that one evidence history serves two destinations, that an external party can inspect it without trusting hosted state, that someone pays, and that the next deployment costs less.
Each assumption is treated as open until an observation closes it.
The native reading. Claim status, evidence burden, proof gates, falsifiers, contradictions, capital-to-proof, governance, downside, and explicit reasons to stop or narrow the thesis.