While I appreciate the fact the amount of thought that went into the OP, the faster units are produced the cheaper it gets to make. Better to take a modest short term loss if it means you can maximize the impact of software while it's still fresh. If consumers hold off too long a lot of new releases will gradually lose value. A well timed price cut means pushing more software out the door upon release, rather than after the novelty of its launch has faded and used copies are cropping up everywhere.
I personally believe they should aim for a holiday price drop, but my sense is that Nintendo will try to ride on new releases between now and Christmas first, then price drop with a Mario Kart bundle next Spring to counter the second wave of Xbox One/PS4 games. For the time being it seems they're banking on hardware sales driven by the rapid deployment of software, some originally intended for "launch window", plus more recently unveiled titles like Tropical Freeze and 3D World.
Also, I disagree $100 drop is necessary to make it worthwhile. I think they'll aim between $50 and $80, and that's enough of a gap as long as it includes the added value of software deeper and more substantial than Nintendo Land. They could also offer incentives that are fairly low risk for them, like download codes rather than physical retail games
It's true that Nintendo limited their options with the way the system was designed, but I don't think the situation is as inflexible as the OP assumes. I also think the analysis ignores some of the risks of trying to sit pretty at the current price point while Sony and Microsoft try to soak up undecided games.