[joesmo]

Suppose that in the example given, changing the existing classes to not be singletons was just as valid as changing the new ones to be singletons. Or there was a good reason for the new classes not being singletons that the code reviewer did not understand. In these cases I'd call the reviewer a code snob. I don't think the reviewer in the example given is a code snob. A code snob to me is one who asks for or makes changes that have no benefit except to himself and might even be detrimental to the project. Since all changes do take time, the benefit of the changes absolutely need to outweigh the time spent making them for them to be valid.