Adriano, row names are really convenient when you want to use them for subscripting. But in this case, you need to consider that you are taking a datetime vector and turning it into text. So you lose all the calendar arithmetic that datetime provides, and you will no longer be able to do things like
T(T.Date > '1-Jan-2016',:)
that you would be able to do if you put the datetime vector into your table as a variable rather than as the row names. It depends on what you want to do.
As Guillaume pointed out, the new-for-R2016b timetable type kind of gives you the best of both worlds, a timetable's "row names" are actually row times, stored as either datetimes or durations.