I may be of help here. You need to solve for other parameters as well, such as ionic strength, water activity as expressed as ionic strength, redox potential and substrate access. I'd first get started on adding a new process-parameter: salt and then expressing its content as ionic-strength. I can provide that data if you cannot find it. This would then modify the information you have already gathered from before: inoculation percentage (and hence starting-dough pH) and temperature, both of which give you a specified generation-time. That generation-time means nothing without the other, directly-correlative outliers sourdough microflora commonly encounter during the course of fermentation.
You'll find that water-activity has little impact on time, barely shaving generation-time by a factor of .0154% at a water-activity of .94 and by a mere .1224% at a water-activity of .99.
All this being said, this model is only predictive for cultures where Lb SF is the dominant microorganism during all fermentative conditions, which is a big assumption. Complicating matters even more, time does not matter in order to understand fermentation. It's merely an outcome.