Classify DO storage-timeout resets as retry-weather

Durable Object storage operation exceeded timeout which caused object to be reset. is Cloudflare's storage-layer failure, not application poison — but today it surfaces as generic errors that pin callers to their full timeouts (observed 2026-07-18: uniform ~160s spec stalls, wakes parked until watchdogs). Tag it into the existing retryable backoff lane the way stream-unavailable and receiver-unavailable are (prior art: apps/os/src/rpc-targets.ts stream-unavailable tagging, PR #1825 poison-skip classification) so wakes and waiters fail fast and retry instead of dwelling on a sick object. Partial mitigation by design — the object stays on its degraded shard — pairs with the slot quarantine task.