That is certainly a good method to maintain epistemic hygiene. All things indeed do need to be double-checked before confirmation into the net. However, would you say you read your spam section and double-check each email as much as you do those in your inbox? Or do you use heuristics to dismiss some things more than you do others? If you do, in fact, expend the same energy on all sources of data then I understand your resistance to the idea of classification by source.