aurthur berman has recently posted an update on the fayetteville shale. you can access it here:

http://petroleumtruthreport.blogspot.com/

scroll down past the haynesville article referenced above.

and with respect to the collapse of natural fractures (and to a lesser extent the propped hydraulic frac's) there was an article published in 1992 by dan stright,etal discussing this phenom in the bakken formation of the williston basin. the title of the paper is:

"Reservoir Characterization of the Bakken Shale From Modeling of Horizontal Well Production Interference Data" SPE 24320.

i dont have a link, but an abstract is available through the spe elibrary.

the generation of hydrocarbons from kerogen results in an increase in volume. in an ultra-low permeability (shale) , pressure builds up and creates an hydraulically induced micro-fracture. when the pressure is depleted, the micro fractures collapse. this has a lot to do with the rapid decline from reservoirs such as the bakken and haynesville.

i have noted previously, that the pressures in the bakken and haynesville is coincidentally about equal to or greater than a normal frac' gradient (~0.7 psi/ ft depth).

By using multi-stage 20X fracing in the Canadian Bakken i.p. rates have nearly doubled. It makes for an upfront first year decline, yet allows for greater amortization of drilling and local pipeline costs. Higher multiple fracing may support a trend for greater intial field production and also lead to an earlier field demise.

i think you have it about right. in the us bakken oil play it has never been demonstrated, imo, that an higher ip rate means greater eur. it may work on the first well(s) drilled in an area as more frac' stages capture more of the natural fractures and micro-fractures, but sooner or later, the wells are competing for the same oil in place. isn't this rule of capture wonderful ?