This is a killer question, because there are no "one size fits all" answers.
5starpix is very correct when he states "Going from 50 and up is much easier then going from 6 and higher", but it's also more profitable to go from #6 up to #1.
If the #50 keyword is #50 because it's insanely competitive, it's best to ignore it completely.
Do you have a good feel for how close you are to #5, #4, #3, #2, and #1? I'd take some time to analyze the pages that are beating me, perhaps using
SEO for Firefox and try to determine if additional investment is likely to be beneficial.
Also, have you looked into the CPC/EPC of a click relating to those keywords? Do they all evince a similar "commercial intent" on the part of the searchers?
Bookmarks