It doesn't make their findings nonsensical because it isn't a condemnation of the paper's gestalt.

While the actual data might suggest that the paper's 'gestalt' isn't wrong (and the difference between negative growth or 2.2%, regardless of the trend, is a big deal for policy decisions made by a country with 90% of GDP in debt), it does reveal that what they presented weren't actually findings; and that's no small issue. Also their response is terrible. Not only do they not mention leaving countries out of the average, they say It is utterly misleading to speak of a 1% growth differential that lasts 10-25 years as small but don't address treating one isolated year of -7.9% as equally large. Finally, they try to get away with saying both It is hard to see how one can interpret these tables and individual country results as showing that public debt overhang over 90% is clearly benign. and By the way, we are very careful in all our papers to speak of “association” and not “causality” since of course our 2009 book THIS TIME IS DIFFERENT showed that debt explodes in the immediate aftermath of financial crises. It's a dishonest and jumbled response to a dishonest presentation of data.

Maybe all economics is like this. I personally doubt it. But if it is, then FTS, it's useless.