What does “Laffer curves are flat” mean for US tax policy?

The following editorial in the Washington Post suggests that raising the top marginal rate would yield lower-than-expected revenues, due to behavioral responses such as income shifting and business structuring.

Washington Post | Editorial Board | Little to gain by raising taxes on the rich

The article appears to mostly be based upon this paper from Rachel Moore, Brandon Pecoraro, and David Splinter at the Joint Committee on Taxation: Laffer Curves are Flat.

Q1. Is this type of model a useful approximation for the typical tax code changes?

From my understanding, this model varies the top tax rate while treating other aspects of the tax code as fixed.
For the purposes of extracting some link between one change and overall federal revenues, that seems reasonable.
But is that applicable to the typical changes made by Congress? Or does the scope of such changes require building bespoke models to estimate effects?

Q2. How do small changes in revenue produce large changes in economic growth?

[WaPo] Though their paper found small changes in revenue from raising the top rate, it also finds significant reductions in economic growth from doing so.
Squeezing those last 0.21 percentage points of revenue out of the rich by raising the top rate to 39 percent reduces long-run GDP by 0.12 percent.
…
In practice, these modest-sounding changes in growth mean millions fewer jobs and an economy that’s worth trillions less, all with less revenue to show for it.