added some info on HUNT cloud deployment

[espenhgn]

No description provided.

[vbabenko97]

I merged current main and moved the HUNT Cloud notes into docs/cloud_demo.md as section 4 (#34 rewrote the README as a landing page, which is what the conflict was), with a one-line pointer from the README's "Adding a TRE" details. While moving it I aligned the steps with the runbook's site scripts:

• scripts/start_site.sh hunt .remote/hunt starts the FLARE client and the mock TRE the executor queries; the original snippet ran ./startup/sub_start.sh from inside startup/, which does not exist there. sub_start.sh --once is kept as the client-only variant.
• the /etc/hosts entry is replaced by start_site.sh's third argument, which patches the server address in a copy of the kit (the project's server is flare-server, not server1, and kits from bootstrap_server.sh dial the public DNS directly).
• kits go through kit_export.sh / kit_import.sh like the other sites; the Drive download stays as the fallback with a note that the archive holds the client's private key.
• the two literal IPs became <DNS> and "HUNT Cloud's egress address", and the firewall sentence now states the direction.

Three things only you can confirm, since I have not run anything on HUNT:

1. Can the lab clone GitHub and pip install the checkout? The kit holds only certificates, so the site needs the repository for the mock TRE and the executor.
2. Is 129.241.176.121 HUNT Cloud's egress address? If so it belongs in the network-opening text as such.
3. Did your run start the mock TRE as well, or only the FLARE client? If only the client, the jobs would have had nothing to query, and the notes should say what was verified.

Happy to adjust if any of this misreads the setup.