I agree that there are MANY sources of error in the basic designs on the web for building such a calorimeter. I'm a retired college professor of human nutrition and am serving as a coach in my first Science Olympiad. My part is to oversee the construction of the calorimeter, and I have very serious doubts about the even remotest accuracy of any results obtained. Some, off the top of my head, sources of error using the pop (soda) can design are heat loss from whatever is holding/suspending the food item; convection heat loss of the air coming from the burning food; radiant and convection heat loss from the can holding the water; heat GAIN from whatever source is igniting the food (matches, propane/butane torch, etc); incomplete combustion of the food item; difficulty/inability to suspend food item on a needle stuck in a cork; inaccuracies in measuring water temperature from uneven mixing. Our team has not yet started building their calorimeter, so there may be even more sources of error identified once we get started! Bottom line is that I do not have even the remotest faith in whatever results we, or any other team, come up with.