Agree, except for that last part.
At this stage OP won't know which backend language/tech will be best suited to whichever idea ends up taking off.
So, best to just use whichever is most popular (and with the best trajectory) since popularity leads to other advantages. Choosing rails now, for example, would not make sense.