Mash pH may vary considerably at the beginning of a mash but will, after 5 or 10 minutes, start to drift assymptotically towards a fixed value which is more or less reached by 25 - 30 minutes in. Whether the prediction of a particular program matches this particular value or not depends on the model the program uses to describe the individual malts. A good malt model says that the protons released or absorbed by a malt in coming to a particular pH is
mEq/kg = a*(pH - pHdi) + b*(pH - pHdi)^2 + c*(pH - pHdi)^3
in which pHdi is the distilled water mash pH and a, b and c are coefficients determined by fitting the titration curve for that malt. Clearly pHdi , a, b and c depend on whether the titration curve was measured at 15 minutes, 20 minutes or 20 minutes so in figuring out how to model a malt we need to decide we are either going to just declare 30 minutes (or 25 or 20) as an adequate descriptor or publish multiple coefficient data sets. In the MBAA paper you have the multiple data set approach was taken and there are some plots which show what kind of variation can be expected for a couple of malts. If we want to get really fancy we can plot, for example, a vs time, fit a curve to that and come up with a function a(t) and similarly for the other parameters. We can then compute the proton deficit/surfeit at any time and come up with a predicted mash pH for 7.5 minutes or 29 minutes if we want.
Clearly none of the available programs are nearly this sophisticated and as their models are pretty sketchy I can't begin to answer the question as to how far into the mash they are trying to predict. As several of them are based on Kai's data I'd suggest looking at his paper to see how long he waited before taking the measurements he did on the malts. I think there may be a fundamental problem there as I think he took the DI mash pH measurement, then added acid or base and then took his single titration curve point measurement.
Another important point is that mash pH may continue to drift measurably beyond 20 minutes. There is usually a pretty substantial pH drop in the kettle and, of course, the first sign of a healthy fermentation is drop in wort pH.
After reading some more threads it seems the Bru'n Water spreadsheet nearly always accurately predicts mash ph, so I'm not gonna worry about this batch and assume I hit 5.3 on the money and I'll calibrate and verify with the PH meter next brew.
