One annecdota: when I was looking for a place to start a phd, the first of my criterias to determine where I liked or not a place was the position of the monitors.
I think that at least 90% of the bioinformaticians I have seen working simply sit down like monkeys in front of their monitor. I just don't understand this. I think that a place where somebody takes the time to correct behaviours like these is a good place to work; therefore, for me in the best place of the world to do bioinformatics, seats are comfortable and monitors are positioned at an ergonomic height.
Other than this, I would like to work for the 1000genomes consortium. I don't know what the conditions and the quality of work are, but there is a lot of data there and it would be good to be the among the first persons to use it.
Another point worth mentioning is whether the position is in a bioinformatics dedicated institute (like EBI for instance) or a more biology-oriented department. I think this makes a big difference for the day to day working atmosphere as well as how diverse the interaction will be.
In my experience, the atmosphere in bioinformatics-only groups tends to be quieter (i.e people essentially sitting in front of their screens), whereas there is much more traffic going on when sharing offices with wet-lab scientists. The type of collaborations that will come up spontaneously over time will thus also be different - bioinformatics-centric in the first case, while much more varied and biology oriented for the latter. Both are essential, of course. A notable point of a more wet-lab oriented environment is that one is more likely to directly interact with the data producers and those that ultimately interpret the results.
Personally, I don't favour any of the above, but it is helpful to know that there might be differences.
I do recommend the Joint Genome Institute for its leadership in genome projects. It's a very well funded place, huge infrastructure, stays in California and you probably used it's data, services or technologies iat least once in the past few months. What do you think?
On the other side of the US, there's Janelia Farm Research Campus, a small and quite modern research facility near Washington DC. People from there work closely to NCBI and Wellcome Trust guys. But, the distance keeps you a little independent. They are fond of innovation and challenges. E. g. Sean Eddy works there and HMMER was developed mostly there (the story is inside HMMER3/Ifernal manuals).
US places pay more and you work less than same level places in Europe. But, Europe is way better to live if you're single/no kids with a good offline life!
In Maastricht the Netherlands ofcourse! Minutes away from three countries and 4 capitals within 3 hours.
No seriously, in my opinion it does not matter where you are located to be a good bioinformatician. We are more and more a network generation of scientists, so it is important to have a good physical and social infrastructure.
When I was orrientating for a proper location, a friend of mine suggested Amsterdam, to finally say that actually it does not matter where you are. Every university delivers good and bad students.
So to answer the question make sure to have both infrastructure available and maybe more importantly, don't be afraid to be mobile by doing interns, fellowships and visits at other institutes.