test262: bump the pinned suite and triage the range

[lahma]

The pinned test262 SuiteGitSha (3655e746...) is weeks old and upstream has moved. Run the standard bump round per the repository conventions: advance the pin to current upstream main, read the commit range for normative changes (git log old..new plus the features.txt diff - a bump is a code change, not a pin change), regenerate, run both configurations, park newly-arrived unimplemented features in ExcludedFeatures, and pay down or document whatever the range turns red - never by editing a test262 test.

[lahma]

Closing as already-satisfied: the premise was wrong - the pinned SuiteGitSha (3655e746) already is tc39/test262 main, verified through three independent channels (git ls-remote, the REST ref, the repo's pushed_at); upstream has had no commit since 2026-08-10. There is no range to triage and nothing to bump.

The round produced a staleness audit of the exclusion table instead, all of it standing:

• All 101 ExcludedFiles entries resolve to real files at the pin; ExcludedFeatures is correctly empty.
• Every externally-blocked entry re-verified behaviorally, not by issue state alone: Add ParserOptions.AllowDirectSuperOutsideMethod adams85/acornima#44/about static data member #46/Add iOS Support #47 all still reproduce (with a probe subtlety worth recording - about static data member #46's forms must be probed through Function(...), the route the test uses; the eval route throws a different parse error and would have wrongly suggested the entry was stale), and the Correct block-decl-func-skip-arguments for the current FunctionDeclarationInstantiation tc39/test262#5112 stale-test entry stands (Jint matches V8/SpiderMonkey).
• The suite is green at the pin: 102,495 passed / 0 failed on a quiet machine. A first run under 4-agent contention produced two 30-second timeouts in staging/sm/Array/toSpliced-dense.js - a file that measures 1.2 s standalone - which is a live demonstration of the existing staging banner's starvation warning, and an argument against ever paying those timeout-bound entries down from standalone timings.
• The "2 need source-phase imports" note in the old context was itself stale: Settle a source-phase dynamic import with the module's [[ModuleSource]] #3074 implemented source-phase dynamic import and those entries are already gone.

Reopen when upstream moves.