Dear Vladimir
Your remark about John Baez and cranks hit home. I had a dispute with John Baez over rejecting my first posting to sci.physics.research inquiring about supernova redshift data and got labeled as a crank in the process. He is the inventor of The Crackpot Index. From Wikipedia: "The method, proposed semi-seriously by mathematical physicist John Baez in 1992, computes an index by responses to a list of 36 questions..."
Petty humiliation can be quite effective. It would be a decade before I posted again. Baez's behaviour was likely symptomatic of a more widespread phenomenon. It seems to me that the situation has gotten worse to the extent that it has become institutionalized, countered for example by VIXRA (thanks to Philip Gibbs) as a reaction.
Here is a little story coincidentally about John Baez's doctoral supervisor. The late Irving Segal (1918-1998) developed a theory he called chronometric cosmology which called for redshift to vary with the square of the distance to account for curvature, instead of linearly as in Hubble's law. He noted that the data collected to support Hubble's law comprised a small part of the data available. Suspecting selection bias, he embarked on a program to analyze as much data as possible. What he found was that the larger data set clearly supported his model, not Hubble's law. This was found over a wide range of wavelengths and with the collaboration of various researchers in a series of papers. In spite of the data supporting his theory, physicists will tell you that supernova data imply accelerating expansion without even considering Segal. I am not optimistic about the future of physics when the critical work of an established and respected physicist is discounted.
Colin