So, first, there is pedagogy and its discontents... we could start a whole other set of threads about teaching methods and strategies, age appropriate and the like. What I object to is self-aggrandizing teachers using questionable data to make grand statements because they want to demonstrate how their mediocre command of an immature subject renders them the universal judges over morality, ethics and the meaning of life.
But then, there is Mendel, the other topic, also complex but more easily managed. His data indeed is unlikely given 95% CIs and Chi-squared tests. Fine. The accusation is that he faked his data. That he made it up and wrote it down.
That is so unlikely as to be libelous.
First, there is the possibility that he just got lucky. After all, historical events happen once and only once. He may have stumbled on unusually good data.
Alternatively, out of the hundreds of experiments he did, he may have thrown out 4-6 that he felt 'just didn't work' for reasons unknown. He may have considered them human error, or not been sure whether the starting seeds were good. Nowdays, we prefer that scientists address all the data good and bad, but those were earlier days. Further, still most scientists tend to discard data they feel is fundamentally flawed. Sometimes entire studies are discarded. We really only get upset when those people are big Pharma or people we hate for other reasons - like ideology and politics.
Personally, I'm inclined to think it was a mix. He discarded a couple outliers and got generally lucky - after all, he was very lucky in his choice of organism, model system, and traits.