fix(terrain): end retries early on a long Retry-After; cover the 429 pause

A Retry-After longer than the 10 s retry budget used to sleep the whole
budget out and then give up, so the proxy's 429 pause started 10 s late
and every new-point poll in that window still went upstream and logged.
The retry loop now stops at once in that case.

Tests: the retry loop no longer sleeps for an unreachable Retry-After;
fully cached points, including stale ones, are served during the pause
with no upstream call; a 503 does not pause; a shorter concurrent
Retry-After never cuts the pause short. CURRENT-STATE now says which
requests are still served and which log.

server/providers/terrain.js

@@ -209,7 +209,8 @@ export function terrainHeightsProxy(options = {}) {
    *
    * The first try keeps its empirically required 30s timeout; network errors,
    * 429, and 5xx receive up to three jittered retries sharing a 10s added-time
-   * budget, with Retry-After honored within that bound.
+   * budget, with Retry-After honored within that bound; a longer Retry-After
+   * ends the retries at once so the 429 pause below starts without delay.
    *
    * A chunk that still fails yields nulls for its own positions rather than
    * rejecting: with several chunks per request, throwing would discard every