I doubt that I am alone here in thinking, "of course I want more data". But I'll just respond on how the data is currently actionable.

The main benefit I see right now is in estimating drug response and metabolism. I have a partially functioning variant of CYP2D6 that makes me a poor metabolizer of CYP2D6 substrates. Something like 25% of commonly used drugs are metabolized by the enzyme coded by CYP2D6, and maybe 10% of the population have one of the non or partially functioning variants.

Also, I'm at heightened risk for a particular fairly prevalent condition which strikes aout 10% of the population, while the estimate is that I have a 30% lifetime risk. A real risk in the general population, and significant one for me.

http://www.snpedia.com/index.php/CYP2D6

SNPedia is a wiki compendium of SNP phenotype information. It has an associated software program (Prometheus) that will process your raw SNP data from the genotyping services and correlate it with SNPedia data, giving a printout of "interesting" SNPs, one's related to medical conditions, and one's related to drug metabolism.

There should be better services like this by now. Anyone know of something?

23andme seems rather pitiful in the information it gives. They map SNPs to clinical disease, but not really to gene function or population prevalence. Prevalence is the main value they could provide to their users, but I suspect they mine that value for themselves, and so don't want to share.

About a week ago, someone posted on the potential for smarties to get their full genome scanned for free.

http://lesswrong.com/r/discussion/lw/7wj/get_genotyped_for_free_if_your_iq_is_high_enough/

https://www.cog-genomics.org/volunteer/