To touch on your 3rd point, that is the inherent problem in these industries: the companies are in it for the long-haul. Do they wind up innovating? No. Epic Systems in healthcare is a fabulous example: http://www.forbes.com/sites/zinamoukheiber/2013/05/15/a-chat...
They code in MUMPS. MUMPS predates C. Yes, you heard me right. The healthcare IT platform that is espoused across the industry is based on what should, for all intents and purposes, be a dead language.
I've taken a bit different approach with some of these industries, and I come from the belief that it's going to require a combination of intellectual partners (like universities and research centers) together with startups to make any sort of meaningful difference.
I'm right now in a niche segment of behavioral health. Funny enough, there is a law that was passed right before the ACA ("Obamacare") that mandates that hospitals put electronic medical records into practice, and over time develop "meaningful use." The big guys have been slow to deliver, the insurance companies want to see change, and the hospitals have to do it for fear of penalty. A lot of people saw disaster, but I saw an opportunity. Over the last several months, I've been working to commercialize some of my research coming out of grad school on this. I was fortunate enough to balance theory and application in school, and this definitely left me jaded towards the SnapChats of the world.
As to your 4th point, that doesn't surprise me at all. I'm Chicago-based, and everyone views the Midwest as a rust belt. I co-led a research initiative that looked at startup activity in the Midwest around ag, health-tech, and advanced manufacturing: http://worldbusinesschicago.com/techcluster
Sorry for the shameless self-promotion, but I feel bad that more dev guys just aren't aware of what's out there. It frustrates me so, so much. We glorify things in the industry that we just don't need.
Here is Siebel's talk: http://ecorner.stanford.edu/authorMaterialInfo.html?mid=2110
Watch it. He's a horribly boring speaker, but the content spoke to me. It's what convinced me to look for the low-hanging fruit in health-tech. Funny; now that I found a problem that can be addressed (that isn't), and have a defensible business model, bringing on a technical co-founder can be so hard for all the reasons we just named. Ha!