Restore Enterprise v1.13.0 content when the features ship

[jstirnaman]

InfluxDB Enterprise v1.13.0 was withdrawn, and #7742 rolled the Enterprise v1
docs back to the v1.12.4 state. Restore that content when the features ship in
an Enterprise v1.13.x release.

What's hidden

#7742 commented out every Enterprise block tagged v1.13.0+ and set
enterprise_influxdb in data/products.yml back to latest: v1.12 and
latest_patches.v1: 1.12.4.

Page	Hidden
configure/config-data-nodes.md	meta-client-* and meta-root-ca mTLS settings; adaptive series-id-set-cache-* settings; cluster, HTTP, and subscriber mTLS settings; user-write-bytes-enabled; OpenTSDB TLS mutual-authentication options
configure/config-meta-nodes.md	https-* mTLS settings
configure/security/enable_tls.md	"Enable mutual TLS (mTLS)" section
tools/influxd-ctl/_index.md	7 mTLS global flags and the mTLS example
tools/influx-cli/_index.md	5 mTLS flags
manage/clusters/replacing-nodes.md	mTLS callout
shared/influxdb-v1/tools/influx-cli/use-influx-cli.md	5 mTLS flags
shared/influxdb-v1/administration/monitor/measurements-internal.md	userwritebytes scoped to OSS with show-in

The Enterprise v1.13.0 release notification in data/notifications.yaml is
commented out, and the v1.13.0 release notes entry carries a withdrawal
callout.

InfluxDB OSS v1 was out of scope and stays at v1.13.0.

How to restore

This repo rebase-merges, so both commits from #7742 land on master
separately. Even so, don't restore by reverting a commit:

• Rollback content spans both commits. 4b1a369ae is mostly unrelated anchor
  repairs, but it also hides one sentence in enable_tls.md that linked to the
  hidden mTLS section. Reverting only the rollback commit leaves it hidden.
• 4b1a369ae also repairs 8 broken anchors unrelated to v1.13.0 — 5 contents
  links on the _internal measurements pages and 3 installation tab links on
  the influx CLI page, on both OSS v1 and Enterprise v1. Reverting it would
  put those back.

Follow the ROLLBACK-v1.13.0 marker comments instead. There are 21, across 8
content files and data/notifications.yaml.

1. Create a branch from an up-to-date master.

2. List every marker:

   grep -rn "ROLLBACK-v1.13.0" content/ data/

3. In each content file, delete the marker's opening comment lines and its
   closing -->, and keep everything between them.

4. In the two content/shared/influxdb-v1/ files, change the escaped
   shortcodes inside those blocks back to live ones: {{%/* show-in ... */%}}
   becomes {{% show-in ... %}}.

5. In content/shared/influxdb-v1/administration/monitor/measurements-internal.md,
   remove the show-in "influxdb/v1" wrapper from the userwritebytes section
   and its contents entry, then restore the Enterprise variant of the
   user-write-bytes-enabled link. The marker there carries the exact line.

6. In data/notifications.yaml, uncomment the
   enterprise-influxdb-v1-13-release notification and delete its
   # ROLLBACK-v1.13.0 header lines.

7. In data/products.yml, set enterprise_influxdb to the shipping version:

   latest: v1.13
   latest_patches:
     v1: 1.13.0

8. In content/enterprise_influxdb/v1/about-the-project/release-notes.md,
   delete the "v1.13.0 has been withdrawn" callout. This one has no marker.

9. If the release ships under a patch number other than 1.13.0, update the
   release-note heading, the v1.13.0+ version badges, and the notification
   text to match.

Leave these alone

The anchor repairs from 4b1a369ae are not rollback content:

• the #tsm1_cache, #tsm1_engine, #tsm1_filestore, #tsm1_wal, and
  #hh_database-enterprise-only contents links in measurements-internal.md
• the [Linux](#), [macOS](#), and [Homebrew](#) tab labels in
  use-influx-cli.md

Check before restoring the release-note details

The hidden v1.13.0 entry says the default replication factor became 2
"(previously 3)". The live v1.12.4 reference already documents 2 as the
default for clusters with four or more data nodes, and that value is a
correction of a long-standing docs error, not v1.13.0 behavior. See
content/enterprise_influxdb/v1/query_language/manage-database.md.

Confirm with engineering what actually changed in v1.13.0 before republishing
the "previously 3" claim. The two statements describe different rules — one
keys off cluster size, the other off creating a database or retention policy
with no explicit REPLICATION clause — so both can be true, but the release
note shouldn't imply the reference page is stale.

Verification

• npx hugo --quiet builds clean.
• grep -rn "ROLLBACK-v1.13.0" content/ data/ returns nothing.
• Confirm the Enterprise influxd-ctl and influx flag tables show the mTLS
  flags, and the Enterprise _internal page lists userwritebytes again.
• Confirm {{< latest-patch >}} output and softwareVersion JSON-LD report
  the shipping version on Enterprise pages.