I think he means that there's a lot of debate when you peel back the covers on how much modern trade deals really impact economic growth. And quite a lot of evidence that they have surprisingly little impact.

If you go read press releases made about trade deals they always have apparently huge numbers but spread over a very large number of years. Modulo what's said elsewhere about the uselessness of economic predictions, the predicted impact is usually very low. Probably because tariffs are already quite low and falling globally, most trade barriers are now regulatory in nature and those tend to be relatively unaffected by trade deals.