Just to clarify this a bit the Semi-emperical mass formula relates to nucleii rather than atoms.
The size of an atom depends on the electron shells that are filled.
It tends to range between .1 and .5 nm
The precise size is difficult to calculate theoretically and there's a lot of quantum mechanics involved.
I think this has been done for simple atoms but only computationally (ie not from first principles) for heavier atoms.
There's something on it here
http://books.google.co.uk/books?id=Fb59En-lF9s C&pg=PA33&lpg=PA33&dq=calculating+electron+orb itals&source=bl&ots=Wc9jtauprJ&sig=5mIew7-QOf7 KVPxtiMh5w2oHOuM&hl=en&ei=zvauSb28O9yxjAepjYyn Bg&sa=X&oi=book_result&resnum=2&ct=result#PPA3 3,M1
There are some interesting side issues. I recall from a question on here some time ago that the reason Mercury is liquid is to do with quantum mechanical effects in the electron shells.