Seriously speaking you are right, this is a risk. I have asked this same question many times to senior folks involved with health reform. I haven't heard a 100% convincing answer, but I'm most of the way there.
Here are the key differences:
1. More MDs, fewer MBAs. HMOs were generally run by insurers (payers) far removed from patients. ACOs are being created primarily by hospital systems and physician groups, sometimes in partnership with a payer (which has more experience managing risk).
2. More/Big data. We have a lot more data, and we've gotten a lot smarter about how to analyze it. So instead of simply rationing care, we can identify and then focus on areas to improve quality and cut costs. Also, the (unsteady) march towards electronic health records (EHRs) should improve this even more.
3. Quality measures. We can leverage data to force providers to compete on quality. See http://www.medicare.gov/hospitalcompare/ for a simple example.
4. Freedom to change plans. In the past, once you had a pre-existing condition you were stuck because no other health plan would accept you. Once plans are prevented from doing this, patients will have the freedom to look elsewhere if they feel their care is being limited.
Update: Added additional info to point 2.
As Bullwinkle said to Rocky, "this time fur sure"
Honestly,
There's little evidence that healthcare is changing for the better and these measures may sound good but in many ways boil down to easter-egging a system broken on much more fundamental levels. You notice the way the mortality rate on the original article's chart was creeping back up, you hear how high-school educated white men experienced a significantly lowered life expectancy in the last ten years?
Of various things in your gilded list, "Quality measures. We can leverage data to force providers to compete on quality". You can use external incentive to increase quality and productivity on something like an automobile assembly line. But turning health care into an assembly line destroys the "human touch" that both helps patients and makes providers self-motivated. There was a study which showed zero gain in health care productivity over some long period of electronic automation. I blame the counter-productiveness of turning things into an assembly-line activity.
Consultants have been slingin' recipes for improved productivity out of managed, controlled, incentivized systems for a long, long time and the disaster has only grown larger. They have made lots of money, they'll continue to make lots of money and transform things for the worse for as long as they can define the problem as being medicine not being run like an automated warehouse. They even make more money failing than warehouse automators make succeeding because they have to reorganize again, and again, and again. Gawd, it's night of the live dead, regulatory-capturing consultants...
The trend right now is moving away from hospital-based care and towards an outpatient-based "medical home" model. In the new model, health systems (not necessarily hospital systems) are given bundled payments for managing a set of patients instead of fixed fees for each exam or procedure. This incentivizes health systems to encourage wellness and rewards them for more aggressively treating the sickest patients and keeping them out of the hospital. This is better for everyone, especially the patients.
It turns out that a small percentage of patients with multiple chronic conditions generate much of the cost of our healthcare spending. One way to keep these patients healthy and out of the hospital is to hire companies that specialize in intensive home-based treatment like the company profiled in the article, HQP.
TL;DR The path forward is keeping patients healthy and out of hospitals, not building new ones.