[roll] Roll fuchsia [sdk] Remove hashes from JSON schemas.

A few of the the jsonschema files shipped in the SDK were
supposedly content-addressed, supposedly to help customers handle
changes to IDK metadata. There's a few problems with this:

- Not all the files were content-addressed.
- The ones that appeared to be content-addressed weren't _actually_
  content-addressed - folks would update them without updating the
  hashes.
- If folks _had_ followed the process and always updated the hashes...
  well, it's not clear how that would have helped.

  It would let you understand exactly which version of the schema some
  bit of metadata used, but you can find that out by just looking at the
  copy of the schema included in the SDK.

  In theory, it could send a clear signal to IDK consumers that
  something had changed, but what would they do with that information?
  If they just crashed, then we could never make any changes (even
  backwards-compatible ones). We could ship both the old and new schemas
  and attempt to get our downstream customers ready for the new schema
  in advance... but if we wanted to do that, we should really ship the
  old and new _data_ as well.

  There was no process that really defined how this would work.
- And to top it all off: nobody really uses these schemas anyway. At
  best, they're documentation, meant for human eyes, not machine eyes.
  But in practice, most IDK consumers will look at the IDK metadata
  itself, rather than the schema.

Anyway, long rant, but this CL deletes the hashes from the filenames.

Original-Bug: 353378507

Original-Reviewed-on: https://fuchsia-review.googlesource.com/c/fuchsia/+/1125713
Original-Revision: cef70aacda6e5bb58fe895f3cd5e48f689fb668e
GitOrigin-RevId: 927a6e20f9152d31e73fe6ea1bc11d3876d93414
Change-Id: If3d85898854be4af070d313a4752416aa414fa28