> I point to issue polling because I'm pointing out the obvious fact that much of what he and his appointees actually done isn't the will of the public
The reason we have elections and not issue polls is because voters balance the totality of the circumstances in reaching their decisions. People may or may not like this or that incident in the moment, but that doesn’t mean they disapprove of the overall arc of the policy. So, for example, polls show that virtually no Trump voters regret their vote and he’d probably win a rematch: https://emersoncollegepolling.com/trump-100-days/ (Emerson is a good pollster, its final 2024 polls nailed Trump’s actual share, and overestimated Harris only by 2). Issue polls taken in the moment also suffer from response bias—e.g. people who approve of the overall arc of Trump’s immigration policy are less likely to respond to polling about a flub than people who are outraged by the policy itself. (The 2024 Harris bounce should be a good lesson in response bias!)
And a rule based on second guessing election results based on issue polling would never be applied even handedly—judges are definitionally elites and such a rule would be applied to favor their preferences. Issue polling showed the ACA under water for quite some time. Roberts had ample basis for finding the statute unconstitutional—far less of a stretch than the expansive interpretations of asylum laws—but he admirably saved it. And I’d argue that, contrary to the issue polls, Obamas’s reelection ultimately bore out that the ACA had a mandate.
> The reason we have elections and not issue polls is because voters balance the totality of the circumstances in reaching their decisions
No, the reason the United States places issues in the hands of a small elite of people elected for fixed terms and some of it in a single individual is a practical feature of the design of a republic which is obviously less democratic than alternatives, like having ballots on most issues or absolutely every implementation detail. It's also a feature designed for practical implementation and less disruptive governance, but so is the separation of powers and rule of law (the "anti-democratic" separation of powers and rule of law having much stronger and more enduring support than actions of any individual president suggests that wasn't a decision the Founding Fathers got wrong...). A rule based on letting the public decide exactly who got to stay in the country and who won each court case would obviously be more democratic than delegating that power to judges, but it would not function particularly well. A rule based on letting ICE elites rather than the judiciary rule on the rights and wrongs of particular immigration cases and giving the president the right to ignore people's constitutional right to due process if they found it inconvenient would not be more democratic and would, I suspect, also not function very well. And suffice to say migration policy is not the only area where Trump is or was arguing that rules shouldn't apply to him or that he should have more power.
Again, the difference between popular sovereignty and populism is quite well illustrated by your argument that if Trump's deportation policy is underwater even in the poll you picked showing him with neutral favourability overall, the "totality of the circumstances" means that deportation decisions he approves and laws he chooses not to follow must be him enacting the will of the people, cheerfully disregarding details like the people explicitly wanting him to do something else and overwhelmingly endorsing the right of institutions ordering him to do something else to make those orders. The belief that the people should have some say in the decision making process is quite different from the belief that the people's will is manifest in every action of some guy they elected on some issue that seemed salient.
The reason we have elections and not issue polls is because voters balance the totality of the circumstances in reaching their decisions. People may or may not like this or that incident in the moment, but that doesn’t mean they disapprove of the overall arc of the policy. So, for example, polls show that virtually no Trump voters regret their vote and he’d probably win a rematch: https://emersoncollegepolling.com/trump-100-days/ (Emerson is a good pollster, its final 2024 polls nailed Trump’s actual share, and overestimated Harris only by 2). Issue polls taken in the moment also suffer from response bias—e.g. people who approve of the overall arc of Trump’s immigration policy are less likely to respond to polling about a flub than people who are outraged by the policy itself. (The 2024 Harris bounce should be a good lesson in response bias!)
And a rule based on second guessing election results based on issue polling would never be applied even handedly—judges are definitionally elites and such a rule would be applied to favor their preferences. Issue polling showed the ACA under water for quite some time. Roberts had ample basis for finding the statute unconstitutional—far less of a stretch than the expansive interpretations of asylum laws—but he admirably saved it. And I’d argue that, contrary to the issue polls, Obamas’s reelection ultimately bore out that the ACA had a mandate.