You're both saying the same thing from different worldviews. Regulations work to minimize risk, but in the process they can and often do also reduce effectiveness and/or progress in a field.

To take a less sensitive example, NASA refused to allow SpaceX to do retropulsive soft landings for their Dragon crew/cargo capsules without a whole slew of regulatory work, which was just not a very effective allocation of resources for SpaceX. So instead of coming down retropulsively, similar to how they now regularly land their massive first stages with, the capsules come down using parachutes and a not-so-soft 'soft' landing as we've been doing since the 60s.

So you could argue this is safer. But that's not clear. What is entirely clear is that the cost and time effort that would be required to abide the regulatory requirements resulted in the entire idea being scrapped, derailing technological progress on that front. Perhaps with some irony, the capsules will still have the thrusters on the capsules to be used in case of an emergency abort - they just can't use them to enable genuinely soft landings.

It's the difference between risk of death, and risk of getting sued. The DIY project has a much lower risk of death, but (if a corporation were to do the same thing) a much higher risk of getting sued.

I disagree those two are decoupled. The DIY project has a much lower risk of getting sued, true. But lower of risk of death is highly dependent on edge cases and corner cases, and I'd challenge you to produce evidence that there is an overall lower risk of death.

None of this is to say I'm against any effort to develop an artificial pancreas. I used to work at Medtronic Diabetes. The slothful pace with which this endeavor is proceeding, despite the mountains of money behind it and the obvious and profound unmet clinical need, pains all of us. But "why was it taken so long?" is a question to which there is no simple answer. And though FDA is very much in the corner of "let's get it done" now, that was not always so.

For instance, the quote from the director at FDA re: "you can do everything but manage diabetes on your phone, you should be able to do that as well" masks the history of the med device industry for years trying to get applications on the iPhone approved by FDA and not being able to do so because you couldn't lock the iPhone down (i.e. prevent programmatically anyone from running any other program).