Like most economic issues, I try not to weigh in unless I have seen empirics/regressions/econometrics and understand the theory behind it. I am not yet to this point on taxes. What I can say is that basing tax rates on personal whimsies or some misplaced moral sentiment is both short-sighted and destined for many unintended consequences. I neither believe inherently in soaking the rich nor defeating some tyrannically vampiric government, though I believe that the income tax is an important and necessary tool at the government's disposal. There are many variables to consider, from the width of the brackets to their height to their multiplicity (though I intuitively prefer an upward-sloping continuous function rather than a piece-wise one for a number of reasons), not to mention various different institutions levying said taxes.
If someone like Beet or phnkrocket or ag want to chime in, I'm all ears, but forgive me if I disregard the notions of non-economists or at least of non-quantitatively-based individuals.
One of the problems with income tax is that it is one of those taxes that hurt the economy the most. After a lots of back and forths on the issue I think that it has been settled that there is a fairly high intertemporal elasticity of labour supply on the macro level (I want to emphasize that I think this to be true, because I suspect I was given a biased perspective on this by my education). Thus, higher taxes on income mean fewer hours worked and thereby lower GDP.
This effect becomes especially marked at high marginal rates, which is what I've been saying in this thread.
There are still plenty of reasons to have income taxes, of course, but it is rather dubious to have them at the high rates some people here seem to want.