We have a way to measure this, it's called the Gini Coefficient, and it spiked in the 80's under deregulation and tax cuts for the wealthy. Since then, it's been out of whack with Western norms and now more closely resembles Latin American wealth distribution.
It would be nice if you shared some extensive numbers to quantify that before I take your word for it. Like total government tax revenue vs upper-percentile incomes and upper-percentile incomes vs actual realized income tax percentages. Like, "here is what the median person making 5M in 1955 paid as an effective rate, this is how many of those people there were as a percentage of the population" and the equivalents for the inflation-adjusted people today. Let's look at some of those deductions too - did a 90% top marginal rate in the 40s cause people to make societally useful donations to get deductions? That could still be very beneficial compared to buying bigger yachts.
Top-end compensation has gone through the roof compared to "regular" worker compensation since we started cutting tax rates on the rich. And we also have no shortage of deductions and shelters today either.
So why shouldn't we raise the rates back to at least reduce the consolidation and government revenue trends that have happened since we lowered them?
Here's one quick source that 5 seconds on Google turned up - https://city-countyobserver.com/did-people-really-pay-91-tax... - estimates suggest that top earners in the 50s payed a 42-45% effective tax rate vs today's 26-28% effective tax rate. That's a pretty giant difference! Interestingly, despite that, the top 1% paid a bit smaller of a fraction of total tax income (30-35% vs 40% today) suggesting that it may have helped spread the money around so that the 2-5%, 5-15%, etc parts of the population were relatively better off compared to the top 1%. Less of the top 1% hitting the top marginal rate suggests that it's a rather useful incentive for keeping your salaries less obscene.
