These articles present an incredibly naive analysis of the situation. The logic is "high pay should equal high stock performance" but this fails to recognize that in an effecient market the ability and salary of the ceo would already be priced into the stock, and thus would have "no effect" on its performance. You can't prove that the stock would've done less-worse if a different (lower paid) ceo was there, which is essentially what these articles try and argue.