So one of the problems is indeed that insurers have conflicting incentives, as you correctly described one of them, and it's possible (although I don't know enough to argue either way) that ACA actually made this worse. I think one of the big arguments for a single-payer system, or at least the "public option," is that those middlemen could (ideally, in theory) embrace the position they're in to fight for lower costs without the conflicting incentives that come with being a for-profit entity. I guess that's basically what Noah Smith (author of the parent's linked article) is saying in his conclusion. There are, of course, other arguments for and against. Healthcare is a fascinatingly (and maddeningly) complex problem.

I agree this is a bad incentive, but a conflict of interest alone does automatically mean an ability to act on it.

In the case of health insurance, driving up the costs requires a monopoly or collaboration of firms without defection. As a result, this means one firm cant act in isolation, and cost increases usually come from the industry not pushing back on outside changes that impact all the firms equally. Examples could be more training for healthcare workers, more expensive standards of care, more regulation and paperwork, ect.

It is pretty telling that UHC prices are not that different than similar plans at Kaiser (which is a vertically integrated non-profit insurance, hospital, pharmacy, and PBM provider). About 15% different when I was picking between them during open enrollment this year.

The incentives go both ways. In the short term, the minimum medical loss ratio rule incentivizes payers to approve more claims and allow network providers to charge higher rates. But longer term payers can only maintain market share with their most important customers (the large self-funded employers) by lowering total medical expenses.