don't overstate empirical claims

[ivoysey]

A nitpick, but: the "popularity" of structure editors is not "evidence" of
their cognitive benefits.  I expect there is something more precise to say
here.