feat(gate): pin reconverge 0.3.0, and prove the gate is unchanged

[vyncint]

The gate ran an analyzer two minor versions old. reconverge has been at 0.3.0 since 22 August; launchbound pinned 0.1.11 in four places — prune.yml, the Action's reconverge-version default, docs/LIMITATIONS.md, and docs/research-baseline.md.

This was deferred rather than forgotten, because a newer analyzer can change what the gate refuses — a change in product behaviour, not a dependency bump. So it is measured rather than assumed.

The measurement

The corpus was run under both analyzer versions, on the same toolchain, and compared:

kernel	clean	caveats	refused	tool errors
histogram	12	0	0	0
matmul-tiled	18	0	0	0
reduce-flip	3	0	8	0
reduce-stable	11	0	0	0
scan-block	4	0	0	0
stencil-1d	45	0	0	0
total	93	0	8	0

The two runs are byte-identical — not merely equal in the totals, but the same candidate hashes, the same REFUSED RC001 lines, the same reasons. The eight refusals are reduce-flip above one warp: the corpus's known flip, and the behaviour the gate exists to produce.

Gate tests pass unchanged under 0.3.0, including known_flip_kernel_disqualifies_above_one_warp and known_stable_kernel_disqualifies_nothing.

No toolchain move was needed

The rule is that the analyzer and the toolchain move together, because a rustc-driver tool must be built by the rustc it wraps. It turns out launchbound and reconverge 0.3.0 already pin the same nightly-2026-04-03 — so the rule is satisfied without moving either. That is what made this a one-variable change and worth doing now.

What this does not establish

General equivalence. reconverge gained multi-warp replay, bounded inlining so a barrier behind a helper is witnessed, and unmasked warp-wrapper analysis between these versions — a kernel exercising those paths could well be decided differently. Six kernels are the evidence, and six kernels are what they are.

That caveat is written into docs/research-baseline.md beside the numbers rather than summarized away as "no change", and docs/LIMITATIONS.md now describes the limits of the analyzer the gate actually runs, with the date it was re-checked.

Verified

• corpus under both versions, diffed: identical
• cargo test --workspace: 41 pass
• gate tests (--ignored): 3 pass
• fmt and clippy: clean

Version bumped to 1.2.0. Tracked in the ecosystem repo as the one open drift item; this closes it.