Honestly I wouldn't put NKS in the same level as TAoCP. NKS is way more experimental, TAoCP is more of a survey of the entire field. NKS is mostly analyzing the results of his experiments, TAoCP is more about the complexity analysis and proofs of the various algorithms.

Also, there couldn't be a bigger contrast between Stephen Wolfram and Donald Knuth. Knuth rightly should have a big ego, but he is a very humble, generous man. Wolfram seems to have a chip on his shoulder, fights with other academics, seems to want to be given credit for being "first!" on things.

I love Mathematica, but Wolfram himself is a controversial figure who often gets in the way of his own work.