You just made an apples-to-oranges comparison. Comparing to the American population, you should be comparing to the county population included in the study. It's given as 636,000 adults. That's about 0.25% of the US. The analysis showed a lot of statistical power.
Actually, no. The population of that county suffered heart attacks at an extremely small rate both before and after the ACA.
There were 377 incidences in the year before - 0.0593% of the county population; vs. 354 incidences 3 years later - 0.0556% of the county population. Controlling for other potential factors, especially region-specific variables and changes in the population cohort over a three-year period, this is not particularly significant.
The greatest differential was among the younger age group (45 to 64) - 102 incidences before vs. 85 three years after - which is the "17%" they are crowing about - but those are vanishingly small numbers in a large population: 0.016% before, vs. 0.013% afterward.
And, even if increased access to Medicare caused this modest improvement, is it really worth bankrupting the country?