Professor Larry Moran is Wrong

I recently encountered another discussion between proponents and opponents of ID (Dr Douglas Axe vs Dr Larry Moran). And guess what, this discussion was again around  protein functionalty and the plausibility of Darwinian paths in the configuration space from one functional protein to another. 

Dr Moran's argument as laid out in his blog Sandwalk is, basically, this. We, stupid IDers, do not understand how it all works. Dr Moran asserts that we cannot say that protein evolution is statistically implausible based on two existing functional proteins A and B whose functions are extremely far apart in the configuration space (in fact, so far apart that a Darwinian path is implausible on the gamut of the whole universe). By the way, there are no provably demonstrable Darwinian paths between any pair in a majority of existing proteins.


Anyway, what is correct, according to Moran, is to hypothesise that the two (sequences of) functional switches:

1. between some protein precursor P and A (i.e. P->A) and
2. between P and B (P->B)

must have taken place in the past (fig.1).

Fig.1. Possible sequences of functional switches between existing proteins A and B and their last hypothetical common precursor P.

Readers are encouraged to decide for themselves, what is more scientific:

• to assess the plausibility of switches between two proteins A and B that are available for measurement now and, based on this analysis, make an educated guess regarding the plausibility of Darwinian paths from their hypothetical precursor P;

or

• to assert that the switches involving some unknown P must have taken place, before we reliably reconstuct the hypothetical paths?

I find it hard to believe that Dr Moran has such a vague idea of what statistics is all about. Can he provably demonstrate a statistically significant difference between the probability of functional switches A<->B and the probability of switches that might have occurred at some point in the past between P and A or B?