@hierarchnoble While I agree with your conclusions on Song (being that enchantments are one of the hardest permanent types to remove if they resolve, or harder to counter), I disagree that all cards must be run for their own face value. I believe this entire deck (and others exist), based on the premise that our opponents will in fact misplay, this deck is a psychological experiment in play style and lines of play. The less our opponent understands whats happening the more likely we will be able to exploit a minor misplay (this deck operates on a maxim of taking stock of many small accumulated advantages over time and converting them into a win).
Just because leading with Bazaar of Baghdad and passing the turn causes the opponent to incorrectly assume my deck choice doesn't mean that the deck is only a psychology experiment. All players try to broadcast misinformation in order to gain a strategic upper hand at all points in the game. That's what bluffing is. I had one of the world's leading Legacy Lands players (Jarvis Yu) Gitaxian Probe me on the play turn 1, see Glacial Chasm, and know exactly what he was up against. I beat him in 3.
What would be a better hypothesis, and one that @POXEVERYTURN has articulated, is that current deck choices aren't strategically equipped to deal with Lands. People get up in arms about how inbred blue is, talking about how maindeck Pyroblasts make no sense otherwise. Blue pilots build for blue and board for Oath, Dredge, and especially Shops. What if they had to build for another deck and simply can't squeeze any more countermeasures into their 75? I believe Lands to be that other deck, and I've had success so far attacking a metagame that is not prepared for a Prison deck that can pivot and combo off with an early Marit Lage.