I've just upgraded to 2.4.0.2 and I've seen a severe change in performance, which I believe I've tracked down to a specific test.
The issue I've seen (and this is a change since the previous version of NCrunch I was using) is that the IDE becomes unresponsive between most operations (typically for 15 to 18 seconds), and periodically crashing. Using ProcExp, I can see that most of the effort is spent in gdi.MeasureString, usually snapshotted whilst waiting for a lock.

The test I believe is responsible for this was parameterized with a *very* long string (>411,000 characters).
[TestCase("POLYGON((296728.92642699997 205150.602408,297355.359789 205352.119332,297355.3597...
public void MakeValidPolygons(string wkt)
{
...
}
I accept that this isn't a normal case, and the performance of my system has returned to normal since I refactored that particular testcase into its own test with non-parameterized data.

However, it's a change of behavior with NCrunch, so I thought I'd bring your attention to it.

John