Ever eager to confirm things for himself, your guide conducted this experiment in his kitchen. In order to protect my microwave, I took the probably unnecessary (but I would recommend doing this to everyone) precaution of placing a half glass of water in the microwave – if there is insufficient material in a microwave, you can blow the internal fuses, rendering the microwave inoperable. However, as the microwave then had to heat the water as well, the melting process took almost two minutes, rather than twenty seconds.
Once I took the chocolate out of the microwave (not have access to Milky Way Stars, I used a block of Ghirardelli semisweet chocolate – a solid block of chocolate conducts heat along its length more than a collection of small chocolates, so I would recommend using some sort of chocolate chips when you do this your self.
Anyway, the measured distance between the melted points from my sample was 6cm.
As my microwave didn’t have a frequency reading on the back, I will use the 2.5GHz “typical” value I found after a brief web search.
Thus: the wavelength is .6m x 2 = 0.12m
Then the speed is 0.12m x 2.5 x 10 9/s = 3 x 10 8m/s, which is a pretty good estimate! If you want to do better, you can try repeating the measurement many times (and making very accurate measurements) and applying statistics to get an average, and an estimate of how much uncertainty you have.