The detail of the paper, rather than the headline,
actually undercuts most of the traditional conspiracy theories. The views of the 90th percentile "elite" are highly correlated with those of the median voter, and largely uncorrelated with business lobby groups.
Paper is at:
https://sci-hub.tf/10.1017/s1537592714001595
Another quirk is that it appears even supermajorities often don't get their way (the graphs show that if 80-90% of the elite favour a policy, it is enacted a little under half the time. Which is better than the probability if 80-90% of the median voters favour it, but still means they usually don't get their way...)
The headline finding is their model which essentially suggests that the median voter doesn't have much influence except where their views align with wealthier people and/or interest groups, but the thing is, they usually do align with one or other and none of the groups actually get their way that often!
It'd also be interesting to see the findings disaggregated by administration and by policy area. It is not difficult to imagine that some of the most consistent disagreements between the median American and the elite American regards issues like upper tax brackets (which elites pay and median Americans don't). To what extent politicians are seeking the favour of 90th percentile Americans when they (sometimes) don't raise upper income tax brackets even when the average American supports it and to what extent they are simply pleasing themselves and following their preferred economic theory are open questions. Disaggregation would also help evaluate other theories, like one of American's political parties being aligned with the elite, or the reverse causation hypothesis where in a political environment where proposed change often fails, the 90th percentile American is [incidentally] happier with the status quo.