I've got a heart issue going on and I've had some fascinating results from capturing my heart rate on a consumer grade chest band monitor, then uploading it to influxdb and viewing it with grafana. Total outlay a couple of hundred bucks. Invaluable to me in tracking what's going on with my heart.
Yet that whole stack would never be approved by... Anyone!
I guess it may be large. Many people have more discretionary free time than money. I surely had when I was a student; I could afford the $50 DIY contraption and assemble it myself, but no way in hell I could find a spare $300.
You pulled 90% out of the air, though. What if it's 42%?
If you think 90% is a better target than 100% for overall public health, well, fine, that's a classic kind of public health policy argument. But without a regulatory system of some kind, you won't know what the actual number is for a given product.
EDIT to add: I'm not arguing that DIY auto-injectors should be illegal to build in general. I'm in favor of people having the freedom to experiment with their own care if they want to.
I am arguing that DIY is not an adequate substitute for the general public unless the actual risks are well understood.
So folks know, there are already at least 2 EpiPen alternatives that are FDA approved.
One is the "Adrenaclick," which is another epinephrine autoinjector that (and this is key) must be specifically prescribed. You can't take an EpiPen prescription and ask for an Adrenaclick, your doctor has to specifically prescribe Adrenaclick. But right now it's a lot cheaper. A friend just switched and his copay went from $200 for EpiPen to $8 for Adrenaclick.
The other is to buy a bottle of epinephrine and use syringes, or buy pre-filled epinephrine syringes. Again, a doctor must specifically prescribe this, but it's a lot cheaper than EpiPen. More complicated to self-administer in an emergency, though--but better than nothing, especially for 3rd parties with access to training, like wilderness guides, teachers, etc.