In your calculation using that table you are leaving out the income of the top 1%. The 99%ile represents the income of the top of "the 99%". In table 2.b, in the main article, you get a taste of the size of the 1% income, which is on average twice the 99%ile.

In the paper I didn't find figures that I could use to reconstruct how much money would have to end up in someone else's hands to achieve the counterfactual, just that it's obvious that the 1% average going from $1,384,000 to $630,000 would represent a lot on its own. The other groups above the median would also have received less of a share. Also note that it may have happened that between 2007 and 2018 the 90-98%iles may have already received some of the money that used to accrue to the 99%ile, who knows why.

I think BI properly quoted the article, but there would be a lot of work to do to contextualize the whole thing. If growth were more shared, would there be more or less growth? That depends on whether the top 1% are actually creating GDP or whether they are just better at capturing the income from it. If there was a system in place that had paid the bottom 98% more, would there have been more inflation, eating away at real GDP growth? What does that big swing from 2007 to 2018 really mean for the 99%ile in C1b, not reflected in C1a?

Honestly, this paper raises more questions to me than it answers.