I guess I can add something. I am (was) a biologist. I did my graduate studies in biology and then phd also in biology (tried to make a switch, but could not find the money). However, for 7 years I haven't done any biology whatsoever. Today I consider myself a CS programmer (c,c++,Perl,R,...) well versed in all three programming styles (procedural, OO and generic). Also, I work as a linux system administrator and a bioinfo "consultant" for small biology groups. But my true interest lies in systemic problems associated to computational complexity and approximation algorithms.
The reason I "turned" was the money. Yes, extremely noble reason. During my studies I figured I could earn some extra cash by doing some programming (first for my friends and then for small bio groups). After a while I found the beauty in slowing problems rather than capitalizing on their end results. Moreover, I think this is ( my personal opinion ) the biggest difference between analytical areas like biology, computational biology and (hard-core) bioinformatics (including areas of computer science and mathematics). While in the former you are putting the emphasis in questions like WHY?? in the latter you are interested in HOW?? (both EQUALLY IMPORTANT!)
I must admit, that making a switch was and still is hard. While biology community looks at you as an outcast, computer scientists don't take you seriously enough. As a result you end up mostly doing freelance-jobs (though I still have a regular job, at least for a year). Switch is not "cheep". You need to invest a lot of time and effort into it if you want to produce competitive solutions and during that time you cannot capitalize on your work (publish papers: since today this is apparently the only measure of accomplishments worthy of career progress in academia (sorry for sounding a little bit frustrated here)), which puts you well behind your peers.
So regarding the question: "Is it worth?" I would say that depends how far are you willing to go. If you decide to stop on analytical fields of computational biology I would definitely recommend it. It will reduce your dependency on others and give you a career push, but be careful, control your curiosity and remember: "curiosity killed the cat" (I realize this is non-scientific but what I am trying to say is that it is important to keep the balance - this is the way of evolution otherwise you might find yourself under a purifying selection)
"Am I just going to decard my past six years experiences ?" "Here Is Something False: You Only Live Once": http://www.smbc-comics.com/?id=2722. :-)
I did wet lab work for 30 years, for the last 5 have done mostly informatics. The biology background does help out occasionally. I find programming to be a bit more rewarding than wet lab experiments on a day to day basis, you don't seem to hit walls as often. (Except when stuff won't install and you can't figure out why - arrgh).
I do make my own sequencing libraries to get a little lab fix and save some cash
How can you have not used statistical methods for years? When I was bench, it was essential
Who says computational biologists are any more or less 'pure' that wet-lab or field biologists?
(I understand what you mean, but don't think you are becoming less of a biologists if you take up a computational position)