The method she uses is sound--but there is a mistake in the problem setup due to unrealistic numbers. The amount of glucose used is in the problem is too large compared to the amount of water present. In reality, a much smaller sample of glucose would have been used in the bomb calorimeter--less than one gram, say.