Use UCX on Narwhal, skip ~/.spack logic for macOS, delete CI workflows on R&D cluster and macOS Intel

[climbfuji]

Summary

1. Provide instructions for loading Cray UCX module instead of OFI module at runtime on Narwhal for better performance.
2. Revert node.js workaround on Narwhal. Requires Bug fix for py-jupyter-server and hidden npm/node dependency spack#417. This also fixed the build error reported on macOS.
3. Skip recent change to ignore ~/.spack on macOS to avoid annoying error messages
4. Delete CI workflows on JCSDA R&D cluster and macOS Intel (both turned off)
5. Bug fix in configs/common/modules_*.yaml: need protobuf and py-protobuf modules for cylc-jupyter server to work

Testing

I tested installing py-jupyter-server@1.21.1 on my macOS with a broken local npm/node installation (to mimick the problem described in #995. I was able to install it successfully with the default variant ~typescript, and also with +typescript (which added npm and node as dependencies).

It also fixed the problem on Narwhal with the existing but broken npm/node installation.

Also: CI testing with default variant ~typescript.

Applications affected

List all known applications (UFS WM, JEDI, SRW, etc.) intentionally or unintentionally affected by this PR.

Systems affected

List all systems intentionally or unintentionally affected by this PR.

Dependencies

• waiting on Bug fix for py-jupyter-server and hidden npm/node dependency spack#417

Issue(s) addressed

Resolves #995
Resolves #928
Resolves #1029
Resolves #1018

Checklist

• This PR addresses one issue/problem/enhancement, or has a very good reason for not doing so.
• These changes have been tested on the affected systems and applications.
• All dependency PRs/issues have been resolved and this PR can be merged.