> Even for rare diseases, a sure thing would be funded.

Only if there is a ROI. Which also means that even treatments for not-so-rare diseases don't get funded if the majority of patients won't be able to afford the treatment at a price that provides that ROI. Clinical trials in the US or Europe are insanely expensive and you cannot even budget them: Adverse reactions by "male patients over 45, born on a new moon Tuesday"? In worst case you restart the trial, excluding the problematic group to at least get the proof of safety and effectiveness for some part of the population so you can finally make some money...

But for an insurance company, the ROI on curing someone’s hemophilia may be very very large.

On the more consumer side of things, being able to fix some sort of obnoxious but perfectly survivable condition that genetic engineering may be able to solve (say lactose intolerant) may be something that you could sell to a lot of people.

From the POV of an insurance company, the patient might go elsewhere in a few years and they would only partially benefit from a cure.

Surely an insurance company would prefer that the patient go elsewhere rather than have to pay $100,000+ every year on treatment. Even if they lose a customer, at least they don’t have that enormous expense.

True. Well, for young people.

For Medicare or certain people on Medicaid it might be worth it.