Apex custom domain stuck in "Validating" with empty errorMessage — DNS + asuid TXT verified, www on same SWA validates fine

### Summary
SWA apex custom domain has been wedged in `Validating` status for 6+ days. Sibling `www` binding on the **same resource** validated cleanly in under 5 minutes and is currently `Ready`. The validator surfaces no error: `errorMessage` is empty/null throughout.

### Resource
- Service: Azure Static Web Apps (Free tier)
- Region: East US 2
- Resource: `swa-usevalentine` (RG `usevalentine-website-rg`)
- Subscription ID: redacted — available privately to MSFT engineers on request
- Apex hostname: `usevalentine.com`
- DNS zone: Azure DNS, zone `usevalentine.com` in RG `platform-prod`, same sub
- Registrar: GoDaddy, NS delegated to `ns1-08.azure-dns.com`, `ns2-08.azure-dns.net`, `ns3-08.azure-dns.org`, `ns4-08.azure-dns.info`

### Reproduction
```
az staticwebapp hostname set \
--name swa-usevalentine --resource-group usevalentine-website-rg \
--hostname usevalentine.com --validation-method dns-txt-token
# add asuid.usevalentine.com TXT matching validationToken, TTL 60
```

### Current state (verified via Cloud Shell against ARM, 2026-05-28 ~02:13 UTC)
- Apex binding status: `Validating`
- `validationToken`: `_52vtb67jw0r4rc18pcgd5yvcudq44d5`
- `asuid.usevalentine.com` TXT (TTL 60) value: `_52vtb67jw0r4rc18pcgd5yvcudq44d5` (matches)
- `errorMessage`: null
- Apex `@` ALIAS A → `swa-usevalentine` (provisioning state Succeeded)
- Activity log: zero `staticSites/customDomains` events on this resource for 48+ hours
- Resource Health endpoint: returns `UnsupportedResourceType` for SWA
- Sibling `www.usevalentine.com` on same SWA: `Ready` (validated within minutes on 2026-05-22)

### What I have already tried (none cleared the wedge)
1. Verified asuid TXT is globally resolvable via 8.8.8.8, 1.1.1.1, 9.9.9.9 — token matches.
2. Confirmed zero CAA records at the zone apex.
3. Deleted the apex ALIAS A, waited 8+ min — no change.
4. Full reset: deleted the binding, deleted asuid TXT, waited 60s for TTL, recreated binding with `dns-txt-token`, fetched fresh token, recreated matching asuid TXT — wedged again for 5+ min.
5. Attempted `--validation-method dns-alias-token` — not a valid CLI option (CLI only accepts `cname-delegation` and `dns-txt-token`). Apex effectively has only one validation path.
6. Re-staged apex ALIAS A first, then rebound with `dns-txt-token` and fresh matching TXT — wedged again for 6+ min.

### History
- First attempt: 2026-05-22 17:51 UTC — stuck Validating ~70 min, no progress.
- Recreated binding: 2026-05-24 07:46 UTC — same wedge.
- Recreated again: 2026-05-28 02:00 UTC — same wedge.

### Hypothesis
Server-side apex validator code path is wedged for this specific SWA resource. www path on the **same resource** validates cleanly, so the validator infrastructure itself is healthy — only the apex branch is broken. Internal validator logs should show what's happening.

### Ask
- Could an SWA engineer poke the validator from the backend, or surface what's causing the silent retry loop?
- Is there a known issue with apex TXT validation right now, or a way to trigger a re-validation that bypasses the wedge?

Happy to share the subscription ID, resource ID, and Cloud Shell logs privately to anyone from the SWA team. Thanks!