> which behaves very interestingly
Check out Bill Thurston's comment on why this is deeply misleading geometrically:
http://mathoverflow.net/questions/53119/volumes-of-n-balls-w...
As several commenters there point out, if you must choose a ratio, a more geometrically natural choice would be the ratio of the unit sphere to its circumscribed cube (which decreases monotonically) or to its inscribed cube (which increases monotonically).