This is a notoriously difficult thing to prove out either way in data, and I’m sure it varies situationally.
The Mariel Boatlift natural experiment did not demonstrate a decrease in wages or increase in unemployment. It makes sense: immigrants both work and consume (i.e., create demand). Unless every immigrant happens to work in the same industry/union, the sum total of immigrants may create demand for labor equal to or greater than they fill.
It also may have the impact you’re suggesting. But it doesn’t have to be zero sum. And, understandably, people only remember when they lost a job potentially tied to immigrant labor. Nobody asks if the job they’re applying to was created due to demand immigrants added to the economy (and how could a company know that?).
Most states have laws restricting faithless electors in some way, including voiding such votes (which has happened). Though, some lack enforcement mechanisms. The Supreme Court has upheld penalties for faithless electors within the past five years. As a result, it’s vanishingly rare.
It’s still a dumb system that is unrepresentative and relies too much on people just doing the right thing, but this characterization isn’t totally accurate.