I've never understood this semantic point of statistics.

When used in English, the word "accept" literally means "fail to reject", and has no more implication than that. Do people making this point use a different definition of "accept"?