It is true that math and nature do not always match up. However, the OP wanted to know how many clutches/eggs it takes to disprove a het. The actual answer is that a breeding test cannot produce 100% certainty that a possible het is not a het. That leads to the question, "How many clutches/eggs are required to get to an acceptable level of probability?" 99% probability is generally considered an acceptable place to stop a breeding test. More eggs simply adds places to the right of the decimal point without ever getting to 100%.
IMO, more people undersell their snakes than is likely to be believed. I once bred a pied ringneck dove from two non-pieds. On one side of the pedigree I had to go back five generations to find a pied ancestor. I just goes to show that a recessive gene can pass down several generations without the human owner knowing it. Even dominant and codominant genes can do this. [A condition (most commonly inherited in an autosomal dominant manner) is said to have complete penetrance if clinical symptoms are present in all individuals who have the disease-causing mutation, and to have reduced or incomplete penetrance if clinical symptoms are not always present in individuals who have the disease-causing mutation. http://ghr.nlm.nih.gov/glossary=penetrance] It is all part of the fog surrounding breeding.