Hmm...first test he's using a heatsink and 2nd test he's not? Or is he just putting thermal paste on one and not the other?
Chances are with no thermal paste it might not make it out of the experiment alive
The thermal paste will have no effect on the performance of the CPU though. It only effects the heat that's produced, which only effects performance once it gets to temperatures high enough to basically cause the CPU to melt down.
As far as thermal throttling goes on the CPU, you might want to do some research on that. Most all the readings I've done on that, the CPU is supposed to overclock itself when it 'knows' that there is a heavy load on it and this is to aid in performance by of course raising the stock speed....However...most findings I came across found out that the CPU doesn't actually detect how much load is on it, but rather the heat. As the heat rises it figures you are putting a heavy load on it and thus it begins to overclock itself.
If he's trying to find out if thermal paste gives you performance, then well I can basically tell you it won't give you different performance results, the CPU will simply have higher Idle and Load temperatures but the throttling will remain the same based on the fact it takes the idle temperature, and when it notices that go up, it'll start overclocking itself (personally I think that's a dumb feature because if the CPU is getting hot the last thing you'd want to do is increase speed and make it hotter)
Perhaps I don't have a complete understanding on what he's trying to do, but I don't think it'll turn out the way you might have hoped assuming I somewhat understand it thoroughly.
Not trying to kill his project, but I know it sucks when you go to do something and it turns out no way in which you originally though.