You say “You admitted above that they added uncorrelated phase errors together. That is a cardinal sin of data analysis”. I don’t know where you learned data analysis, but this feels to me like a principle you just made up in order to criticize Bouman. She’s got an algorithm that (among other things) does something whose effect is to add up some phase errors. It happens that doing so makes them cancel out. There is nothing whatsoever wrong with this. (The errors are systematically related to one another, though I don’t think correlation is the right language to use to describe this relationship.)
You say “You admitted above that they added uncorrelated phase errors together. That is a cardinal sin of data analysis”. I don’t know where you learned data analysis, but this feels to me like a principle you just made up in order to criticize Bouman. She’s got an algorithm that (among other things) does something whose effect is to add up some phase errors. It happens that doing so makes them cancel out. There is nothing whatsoever wrong with this. (The errors are systematically related to one another, though I don’t think correlation is the right language to use to describe this relationship.)