Second, it only makes sense that the easiest paths (timing and avoidance) would be the first paths taken..and thus show larger impact when marginal rates change.
3rd, this is the same guy that knows (as he wrote a paper on it) research on levels of capital investment based on marginal tax rates, in his language, shows that, in his own words, Atlas does indeed shrug.
Their research examines the income tax returns of a sample of sole proprietors before and after the Tax Reform Act of 1986 in order to determine how the substantial reductions in marginal tax rates for the relatively affluent associated with that law affected their decisions to invest in physical capital. They find that individual income taxes had a large negative effect, implying that a 5 percentage point increase in marginal tax rates would reduce the proportion of entrepreneurs who make capital investment and mean investment expenditures by approximately 10 percent. In their words, these particular Atlases do indeed shrug.
http://www.bus.umich...introfinal899.pdf
The interdependence cannot be ignored. I will posit, though, that the most prudent action of the Treasury would be to not touch income tax rates AT ALL, and instead focus on eliminating paths of evasion and avoidance (loopholes) first. The net effect will be positive on revenue and will not incur the kind of explosive noise associated with legislating the tax tables. It would probably also quiet the guys like Warren Buffet, who pay people to avoid taxes for him and then complain that his secretary pays more ;-)