Rubian and Kestowv public benchmark standard

Public results must identify:
- the component under test
- the work completed
- the control or comparison class
- the correctness condition
- the observed result
- the relevant tradeoffs and limitations
- whether the result is a single observation, peak, median, or repeated estimate

Public results must not expose:
- internal hostnames, addresses, paths, or account names
- source commits, private receipt identifiers, or raw output
- proprietary mathematics, scheduling models, topology, or implementation details
- private transport or connection architecture

Detailed evidence is retained for controlled review. A public summary is released only after infrastructure, IP, correctness, and claim review.

No Rubian or Kestowv result automatically becomes a Project Valux result. Project Valux claims require a new matched test of the relevant system path.