Further emphasise filesystem backups don't work (#67634)

In #33102 we added a warning against using filesystem backups.
Experience has shown that the wording we added was insufficiently
general and open to misinterpretation. This commit reworks it to be
clearer.

This commit also clarifies that snapshots are not incremental across
repositories.

Author: DaveCTurner

docs/reference/snapshot-restore/index.asciidoc

@@ -18,14 +18,16 @@ Snapshots can be stored in either local or remote repositories.
 Remote repositories can reside on Amazon S3, HDFS, Microsoft Azure,
 Google Cloud Storage,
 and other platforms supported by a {plugins}/repository.html[repository plugin].
+// end::snapshot-intro[]
 
 {es} takes snapshots incrementally: the snapshotting process only copies data
 to the repository that was not already copied there by an earlier snapshot,
 avoiding unnecessary duplication of work or storage space. This means you can
-safely take snapshots very frequently with minimal overhead. However, snapshots
-are also logically independent: deleting a snapshot does not affect the
+safely take snapshots very frequently with minimal overhead. This
+incrementality only applies within a single repository because no data is
+shared between repositories. Snapshots are also logically independent from each
+other, even within a single repository: deleting a snapshot does not affect the
 integrity of any other snapshot.
-// end::snapshot-intro[]
 
 // tag::restore-intro[]
 You can <<snapshots-restore-snapshot,restore snapshots>> to a running cluster, which includes all data streams and indices in the snapshot
@@ -39,17 +41,22 @@ You can use
 to automatically take and manage snapshots.
 
 // tag::backup-warning[]
-WARNING: You cannot back up an {es} cluster by simply copying
-the data directories of all of its nodes. {es} may be making changes to
-the contents of its data directories while it is running; copying its data
-directories cannot be expected to capture a consistent picture of their contents.
-If you try to restore a cluster from such a backup, it may fail and report
-corruption and/or missing files. Alternatively, it may appear to have succeeded
-though it silently lost some of its data. The only reliable way to back up a
-cluster is by using the snapshot and restore functionality.
-
+WARNING: **The only reliable and supported way to back up a cluster is by
+taking a snapshot**. You cannot back up an {es} cluster by making copies of the
+data directories of its nodes. There are no supported methods to restore any
+data from a filesystem-level backup. If you try to restore a cluster from such
+a backup, it may fail with reports of corruption or missing files or other data
+inconsistencies, or it may appear to have succeeded having silently lost some
+of your data.
 // end::backup-warning[]
 
+A copy of the data directories of a cluster's nodes does not work as a backup
+because it is not a consistent representation of their contents at a single
+point in time. You cannot fix this by shutting down nodes while making the
+copies, nor by taking atomic filesystem-level snapshots, because {es} has
+consistency requirements that span the whole cluster. You must use the built-in
+snapshot functionality for cluster backups.
+
 [discrete]
 [[snapshot-restore-version-compatibility]]
 === Version compatibility
@@ -116,4 +123,3 @@ include::monitor-snapshot-restore.asciidoc[]
 include::delete-snapshot.asciidoc[]
 include::../slm/index.asciidoc[]
 include::../searchable-snapshots/index.asciidoc[]
-