People dying because they can't afford insulin are usually not dying because they have insurance where the prescription co-pays for insulin is too high, but because they either are actually uninsured, are effectively uninsured because of a high deductible, or have insurance that works on a coinsurance rather than co-payment model.
And even then that's usually not really the problem, the problem is usually that they can't afford a more basic life necessity (often food or shelter) for themselves or their dependents, and voluntarily forego (or limit use of) insulin to afford the other need. Which is a real problem (and a real public economic policy problem), but usually not fundamentally a healthcare policy problem.
(On the other hand, it's often a problem where moral hazard arguments of the type at issue here contribute to the failure to solve, so the argument about that being the problem may still be correct by a less direct route then seems to have been intended.)