Scenario: you want part of the page to load (e.g. all content from the main origin), take a snapshot then let the rest of the page load and take another snapshot.

It's possible today to write a protocol handler to arbitrarily delay fetches, however the virtual time code knows nothing about this.  We need a way for it to know which resource requests matter.

This is challenging because the resource filtering is (probably) done browser side but the virtual time code is in the renderer.