This is interesting in that firstly, it presupposes that the standard checks for water (drains - and on some 172s an awful lot of them) are not 100% succesful. Does anybody have experience that is the case ?
I have never read it in an accident report (I ain't perfect of course) other than on one occasion when there was so much water it completely filled the fuel tester and thus with a single amorphous liquid fooled the guy testing into thinking it was fuel.
Secondly, as a principle it can be applied differently to different aeroplanes. A PA-28 can only be run on left or right. A 172 (for example) can be right / left or both.
Typically I have been running the 172 on both right through taxi, test and tak-off, however (as an engineer) I am well aware that does not mean (head losses in hydraulic systems etc.) that the engine is always drawing half its demand from each tank. Thus if my "firstly" can give rise to a problem, the manifestation of the problem would depend on when the engine started to draw from the tank with water in it.
If the aeroplane had been left with the selector on both and was across a slight slope, that could well give a gallon or two more in one side that the other. If it was the lesser side with the water in, it could well be quite some time until that water got to the important bits. Murphy says that would be at a Bad Time.
Hmmmm. does this give cause for concern ? Discuss.