Why financial innovation is bad

Is it just me, or does this NPR Planet Money podcast completely miss the point. Felix Salmon makes two excellent arguments:

1. We need to control leverage (ie. impose capital controls)
2. Most financial "innovation" has been about escaping capital controls and taking on hidden leverage.

Tyler Cowen, who takes the other side (maybe just for the attention) makes the usual glibertarian pro-market arguments.

First, Tyler does not seem to understand the difference between equity financing (which restructures balance sheets, but does not expand them) and debt financing (which expands balance sheets). The former is pretty innocuous, the latter is potentially cataclysmic. Felix does not press this point, maybe he doesn't understand quite how important it is. Tyler bringing up venture capital is egregious.

Second, neither of them seem to understand that the liability side of bank balance sheets is no place for market discipline. We have both empirical (last 75 years) and theoretic evidence for this (DD equilibrium).

Third, debt financing in particular should be focused on credit analysis, so innovation should take the form of better credit analysis. It isn't hard to get the incentives right for this -- just mandate that the loan be kept on the lender's books.

Fourth, the empirics behind the benefits of securitization are awful. Take the original MBS with passthrough -- it began in 1970. Did the US exhibit ANY inability to make enough houses to shelter it's population pre-1970? Since it did not, what real, practical purpose did MBS serve?