Excellent post, you said it far better than I did.
Can't help but notice that a few days ago people were generally agreeing that anything beyond about t+72 was highly unreliable, yet as soon as anything particularly cold and/or snowy appears on the charts at t+96 onwards people once again get sucked in. Likewise if, in the same FI timeframe, the model output then changes to something less favourable.
Let's have a little more analysis and objectivity here and stop taking much notice of anything past 72 hours.