For what its worth, here's a link to a 2008 article in Model Railroader about foam shrinkage (or lack thereof): http://cs.trains.com/mrr/f/11/t/139257.aspx
OTOH, just to confuse things slightly, the construction industry does report some shrinkage, to wit, here: https://www.buildingscience.co...rinks_Correction.pdf but by my read, this is largely thermal expansion/contraction, as opposed to 'aging' under a fairly stable thermal environment.
I suppose one could look at what DOW and/or Owens Corning have to say on their websites...