If you want to find out who out of two groups decides, you can't look at the cases where they agree.
People at the 90th income percentile are not "a group" and don't make decisions as a group. They are merely a population of individuals that these researchers decided to single out.
If, for example a 30000 people group always wins when they disagree with a 30 milliion group, then either the 30000 group decides, or its interests align perfectly with a third group that does decide.
The paper shows a correlation involving people around the 90th percentile and people around the 50th percentile. The paper simply provides no data over who decides in cases where "the 30000" and "the 30 million" disagree. So you cannot infer that "the 30000 always win" from the paper or any of its references.
It doesn't ignore them. They enter into the the calculation, but they don't provide any information about the power structure.
"Domination" and "power" are quantitative issues. If on most issues, governmental decisions reflect the will of about 50% of the population, then you can't say that a small group "dominates". Even if the correlations they found were valid, they would amount to nothing more than "slightly disproportionate influence", not "domination". And that the 90th percentile has a slightly disproportionate influence on politics and policy is no secret and pretty uncontroversial: that population consists mostly of older people and more politically active people.
You speak as if the economic mobility in the USA were perfect. it isn't.
This has nothing to do with economic mobility, it has to do with demographics. Many people reach the 90th percentile after working for a few decades and drop out again when they retire. They certainly aren't in the 90th percentile while getting their education.
I don't understand your reasoning here. The groups were selected solely based on their income. They weren't selected based on their opinions or how closely they correlate with political decisions (that would be circular). The groups don't need to have consistent or stable preferences.
You are confusing a sociological and political meaning of "groups" with a statistical meaning of "populations". People whose last name starts with the letter "X" are a population, but they aren't a "group" in the sociological or political sense.
Why? I don't think that's obvious. You're saying that if I to a multivariate analysis of the preferences of a large group and a small group vs. political decisions, I should expect to find most of the explanatory power with the small group, regardless of which small group that is? That's a pretty outrageous statement, which needs to be substantiated.
Let me try to illustrate. Assume the US consists of 100 subpopulations, each with 100 preferences they particularly care about, but overall, the population is divided equally on all preferences. If you randomly pick one person from each of those subpopulations and look at how policy aligns with their preferences, you won't see any consistent agreement because you just picked 100 random people; any preference that you identify within that population is just a statistical fluke. If you pick 100 people out of just one (or a few) of these populations, they do have many consistent preferences. And since politicians do implement something, you'll actually see some of these agree with policy. It's not "outrageous" at all. More importantly, it's the authors' job to show that their statistical methods are reasonable and valid, which they haven't done at all; there are no control groups in the paper, for example.
If you have a good dictator, nothing beats a benevolent dictatorship. The problem is that dictators tend to go bad. When those that wield power are out of touch with the general population, they tend to use it to further their own interests at the expense of the people. Democracy is the best safeguard against that that has been found.
Democracy by itself doesn't deliver that. It is the combination of democracy and limits on the power of government that protects our society from dictatorship. But limits on the power of government mean that many preferences that the population has simply shouldn't be heeded by government at all. Most of the preferences of the 50th percentile in this study that representatives fail to implement at the federal level are probably preferences that should be ignored by the federal government in the first place because heeding them is not the function of government in a free society.
As I was saying: the fundamental assumption that we ought to make decisions as a "majoritarian electoral democracy" is wrong; that is not what our nation is supposed to be, and for good reason. That's one of the reasons that even if the correlations they observe in the paper exist, there would be nothing wrong with it. It is not intrinsically anti-democratic, it is simply not a "majoritarian electoral democracy" because it is not supposed to be.
The terminology of the paper is confused because we actually do have "majoritarian electoral representation", but the paper tests for simple "majoritarianism" on each issue. The former is not intended to deliver the latter.
If people want to argue that we ought to change our political system to that, we need to have an argument about the fundamentals of our society, not about hokey statistical analyses.