What should the OpenUSD stack-maturity ledger validate next?

[nicolaspopravka]

The benchmark now preserves a 15-result OpenUSD build-method benchmark plus
published Cycles and MoonRay delegate results. The next useful benchmark run
should be selected by how much it can show and how reproducible it
is, not by whether it is likely to succeed.

Use this thread to suggest:

• an exact VFX Platform year, OpenUSD build method, delegate, and version;
• the USD scene or asset and which part of the stack it exercises;
• the intended runtime, graphics backend, hardware, and driver;
• a related upstream issue, pull request, or maintainer question, if one
  exists;
• whether you have access to relevant hardware or a licensed renderer; and
• whether you can review the proposed scope or resulting outputs.

ASWF reference, Pixar control, and delegate diagnostic runs remain separate.
Partial runs, failures, missing images, and visual regressions are publishable
findings when the run's outputs are preserved and attributable. Timings from different
hardware or drivers are not performance rankings.

This first phase is for feedback and coordination. It does not accept external
code or benchmark artifact submissions, promise paid compute, or authorize a
new renderer run. Concrete proposals will become maintainer-run issues only
after scope, licensing, cost, output requirements, and responsibility are agreed.

Current orientation:

• The OpenUSD build-method benchmark is published for CY2023-CY2027.
• Cycles 5.2/OpenUSD 26.03 has a transparent-output investigation with a
  unmodified-source cross-version control pending.
• MoonRay XPU is not built and will not proceed until build review and
  NVIDIA validation are approved.
• Modern commercial-delegate coverage is not scheduled.