Allow for AI Gateway timeout value to be set on GitLab Dedicated instances

Proposal

Currently, it's possible to change the AI Gateway timeout only on a self-managed instance - with the setting being hidden on Dedicated instances.

With this setting being unavailable for Dedicated customers, it can cause issues with some prompts/LLM requests. (example - internal link)

It should be possible to increase the AI Gateway timeout for Dedicated instances to avoid problems.