You are right in saying that if we don't have better evaluation methods, it doesn't matter too much. The trouble is, should we just ignore the problems with cur

You are right in saying that if we don't have better evaluation methods, it doesn't matter too much. The trouble is, should we just ignore the problems with current methods? It seems the equivalent of sticking your fingers in your ears and saying, "la la, I can't hear you."