Anyway I would like to briefly mention importance of the video recording. It can help solving the tedious problem of performance time recording so that participating designer can focus more on the qualitative cues like "facial expression, mouse searching patter and untold problem the user may have. For our objective is improving the design, a short clip can help persuading stakeholder or developer who thinks this state is good enough. It is already working, why should I change this! Showing a repeated pattern of frustrated user struggling is a good method to change this thinking. It is a good practice marking the video so that I can refer it later, if the software package permits. But even if you can't mark the recording, leaving a short memo about this ,will remind you to search the video to see it again.
One thing I would like to point is that the suggested evaluation metric like performance time, number of error, user surveys are usually dealing with the usability part of the interaction design. However human are not only logical but emotional. And as you can remember in the paradigm of the design, emotional part like aesthetics, pleasant to use should be an important design paradigm. If you can't measure something, you can't improve it. So how we can evaluate emotional performance is getting more important. I am expecting to see "Web Experience Analysis" by Dr. Mihaela Vorvoreanu to solve such a problem.