You can be pretty sure that statements like this, which sound beautifully exact and convenient, have little basis in fact. I suspect they are typically the result of people either just making something up or extracting a sound byte from a study or comment with some merit and spewing it out of context as a new rule of thumb.
You can be sure there's no way that "rule" will hold up to much scrutiny without first qualifying it to death. What shape is the athlete/drinker in? What age/sex are they? Is this specific to swimming or asserted as true for all disciplines? What sort of training session is it - speed work, strength work, endurance, recovery session? Do 2 drinks cost one training day if it's adjacent to the drinks and as a result of training while still effected or is it that the physiological harm caused by 2 drinks is supposed is simply equivalent to the benefit of one "training day". Is this two drinks back to back or any two drinks even if consumed half a drink at a time over a four week period?

As you've posted the statement, presumably as this guy related it, it's pretty meaningless.
Whether or not it's derived from a study with some validity is another matter entirely but you'd need to know the details for it to have any meaning to you.
P.S.
Alcohol definitely impacts your CV system and I've occasionally gone for a run or cycle the morning after several drinks and noticed an increase in HR for a given effort of 5-10bpm. As I understand it alcohol causes blood vessels to constrict temporarily. But attempting to quantifying the impact with a general catch all statement like this is bull.