Elik,

RTC has streams, we don't use the word branch and I think there are differences. Users work against a stream and deliver their work there. Using a flow target to multiple streams allows you to integrate work between streams.

You can create a new stream from another (call it branch, if you like) or from a snapshot, or from baselined components.
The new Stream is the one you want to use for the new stuff. The old one you want to use for maintenance.

You can develop independently against each stream. If you have changes on the original stream, that you want to promote to the new stream, you point your repository workspace for the new stream to the old stream. You add the old stream as flow target an make it current.

Then you see the incoming (and outgoing) changes - the differences between your workspace and the flow target. You can now accept (and merge in) the changes on the old stream into your repository workspace.

Now, without delivering to the old stream, change your flow target to the new stream, integrate and test the changes there. Finally deliver to the new stream. This brings the integration of the new stuff on the old stream and the changes on the new one back into the new stream.

This is, in theory, how this is supposed to work.