There is no real‑world policy outcome that directly matches Chamath’s hypothetical, so his quantitative claim can’t be cleanly scored as right or wrong.

Key facts:

• In March 2021, Congress passed and President Biden signed the American Rescue Plan Act (ARPA), a $1.9 trillion COVID relief package, largely as proposed rather than a smaller, highly means‑tested alternative.(en.wikipedia.org) Chamath’s scenario was explicitly instead of that $1.9T package.
• ARPA combined many elements: $1,400 recovery rebates, extended unemployment benefits, large state and local aid, school funding, health subsidies, expanded child tax credit, etc. It was not structured as a purely means‑tested transfer heavily concentrated on the poorest and excluding the richest, which is what his statement envisioned.(en.wikipedia.org)
• Analytical estimates suggest that some components of ARPA line up in magnitude with the range he mentioned. For example, the Penn Wharton Budget Model estimated the $1,400 “recovery rebates” alone would cost about $416–465 billion, depending on the version,(budgetmodel.wharton.upenn.edu) and it estimated about $120 billion for the expanded child tax credit and roughly $350 billion for enhanced unemployment benefits within the broader Biden relief plan.(budgetmodel.wharton.upenn.edu) But deciding which combination of components corresponds to his hypothetical “really expansive” means‑tested program is subjective.

Because:

• The specific means‑tested‑only package he described was never enacted, and
• Any attempt to back‑fit his $500–700B estimate to pieces of ARPA relies on judgment rather than a clear empirical test,

we cannot definitively say his prediction was empirically right or wrong. It is therefore best classified as ambiguous rather than correct, incorrect, or merely “too early to tell.”