The problem is probably that you have to train in two separate fields to be competent in computer science and biology, and that's not going to be common.
But the quality and sophistication of software in, say, astronomy makes bioinformatics software look like it was written by 14 year old spammers in the former soviet bloc. The real problem is that biology went from being essentially a liberal art, more like history or sociology than a science, to an extremely quantitative field in only a few academic generations. You can get a PhD in Molecular & Cell Biology from Berkeley without ever having taken a statistics, linear algebra, multivar calc, or CS course. Biologists need programmers more than the other way around, so they should start by taking a few courses with actual numbers and stuff and learn the bare basics first.
