Papers aren't published under real names for the sake of the papers or to increase the acceptance of the claims in them. They're published under real names for the sake of the authors writing them. "See, I did that!"
It would be hard to evaluate a researcher's abilities if one couldn't read her papers.
This is exactly my point. When assessing research, the evaluation of a researcher should always be far, far secondary to the evaluation of the research itself.
In the contrary case, which is a separate issue - that of hiring a scientist, in every other arena, employees are hired based on interviews and testing, with some consideration given to past experience - but this is more and more, especially on the cutting edge, the least pertinent concern.
In science however, employees are hired sometimes entirely based upon their "experience", with little to no consideration given to their presently-assessable ability.
If I were to hire a research scientist, I would far prefer to test potential suitors by having them walk through the conducting of research in front of me. Many, however, would balk - indignant, and stand proud on their legacy of papers instead.
This gives me no idea of their actual ability, and taken on face-value - ignoring tradition and dogma - is actually a red-flag.
In principle, why should we care who makes a particular claim? The claim should be decided on its own merits.