This is an "out of bound" question I must say since it is a factual question..
It is fine that multitape TM and single tape TM has same expressive power..
But as far as computational complexity is concerned(quoted from wiki) :
This model intuitively seems much more powerful than the single-tape model, but any multi-tape machine, no matter how many tapes, can be simulated by a single-tape machine using only quadratically more computation time
Reference : Papadimitriou, Christos (1994). Computational Complexity. Addison-Wesley. p. 53. ISBN 0-201-53082-1.
Hence A) is the correct answer..