Improved UX for pushing data to WARP server

Currently outlined steps for pushing to WARP server:

1. Enable network WARP in settings
2. Add API key in settings
3. Open a binary and annotate
4. Create WARP file with selected annotations or all
5. Use the commit file command or upload via website

This is all documented and fine but making the user have to create the file and then commit is unnecessary friction in most cases.

We should have a "fast path" here that will "just push the information" similar to other products. Optionally we should also give the option for the user to automatically push information at certain key points, such as when analysis completes. Because of this automatic pushing we should be providing a "default source" setting for users to set (should be settable from the WARP container tab).

On the server we should be more aggressive in "tombstoning" functions after push operations to avoid duplicate function information poisoning a specific file (e.g. the one which had multiple pushes occur on the same function information)

Start by tracing the existing WARP settings, WARP container tab, commit file command, and website upload flow described in the issue. Determine where a direct push, optional automatic push points, and a default source setting would fit, then review the server-side push handling. Done means the fast path and requested automatic behavior are defined without duplicate function information for a specific file.