This data isn't of much use for analysis without all the startups that failed or didn't get acquired. This is like trying to analyze average height by nationality using a sample consisting of professional basketball players for each country.
Now the question is - is being acquired a good trace marker for a not-particularly successful startup? The answer is it doesn't matter - because the original PG essay was talking about being a major economic driver, so in some ways it is irrelevant what the non-successful startups are doing.
What it comes down is, is acquisition a fair sampler of successful startups. I don't know the answer to that, but I think a fair number of people would be surprised if it was "Hell, no".
Back to the sporting analogy, it is like trying to figure out the fastest nationality by looking at their 100m record. If each country trains and sends to the Olympics its fastest people then yes - but you have no simple way of knowing whether that is true. However you may draw some conclusions about who the fastest race is, on the basis that you would expect any bias to work against them.