You're making too many gross generalisations here, as does Mr Taleb. He's right that in all fields there are statistically illiterate people making huge errors, and academic theorists have been making such - and more complete - criticisms for a while now. So, there's a real issue. However, there are a good number of competent social scientists too, beyond the mediocre morass, who understand just what variable construction and limitations to statistical operations are.