The post is on a funda I call 3CR i.e. confirm, Confirm, CONFIRM, Report. Basically, It says you can easily make a ass out of yourself if you report premature findings with no solid proof quite easily. Hence, double..no..triple check everything before establishing or generalising something. This has happened to me multiple times and hence I push myself to adhere to this principle as much as possible(quite hard at times for me)
Ofcourse the above becomes more important when you are reporting to your boss or client where reporting any findings which may turn out to be wrong could be quite a disaster.
So, before jumping from your seat and yelling Eureka, just run a few more testcases ;)