rshe
If you're considering a startup, ask yourself: am I passionate about what this startup is trying to accomplish? If you really are, join the team. Get on the bandwagon. You get equity, which means your successes are shared. If it's just another startup that you don't care about, it's not a good match anyways.
I do think there are many great papers coming out in biology today, and scientists are still fully capable of writing insightful books and essays for the general public. I can see why some papers feel like a collection of trivial data, but trust me, beautiful and convincing data is well appreciated. While exaggeration of results is also a problem, we are trained to read all papers with a critical eye. There are always good papers and bad, but here are some links to ones that I think are good:
http://www.nature.com/nature/journal/v397/n6715/full/397168a... http://www.nature.com/nature/journal/v437/n7058/abs/nature03... http://www.cell.com/abstract/S0092-8674(09)00963-5 http://www.sciencemag.org/content/324/5928/807.abstract (Hopefully you'll be able to access these - if not, that's a whole nother problem about academic papers)
I won't comment too much on your generalizations, but I want to note that it is hard to predict a priori which findings from academic research will become useful for industry later on. I think you'll find defunding academic biology to be a pretty unpopular viewpoint. Perhaps you could elaborate on what serious biological research means? (Plus, I'd say paying graduate students 30k/yr is a pretty efficient labor force)
On a more macroscopic scale, talks also allow scientists to highlight deeper themes that are often lost in the minutiae of a technical paper. This is especially important in biology because we want to find universal paradigms from experiments done on model organisms. A talented speaker can distill the most important themes from a body of research in a way that writing rarely achieves.
In summary, talks are a great medium for conveying conceptual narratives. In biology talks, the important assertions are almost always backed up by a slide that shows real data. However, if I am an expert in a particular subfield and really want to get into the details, of course I'll go read the paper.