Seems like this article missed its due attention here on coffeed. http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6T6R-4W2NDVW-5&_user=8771778&_rdoc=1&_fmt=&_orig=search&_sort=d&_docanchor=&view=c&_searchStrId=1128188201&_rerunOrigin=google&_acct=C000050221&_version=1&_urlVersion=0&_userid=8771778&md5=ebc313cd75988ded66e8324dc0b0f9df
It took me three or four reads to finally understand what is being said here. If I've got it right, the abstract is that in untreated water without significant Na+ a key determinant of water flow rate (or brew time if you prefer) is the carbonate/bicarbonate content on the brew water. This is because the bulk of the bicarbonate (HCO3-) breaks down to carbonate (CO3-2). Carbonate then breaks down into dissolved carbonic acid (CO2) during extraction. The author indicates that this release of CO2 as gas contributes to coffee bed compaction and therefor reduced flow rate.
In softened water (where calcium and magnesium ions are replaced with sodium ions), the Na+ content is responsible for the significant impact on flow rate as it reacts with bicarbonate to yield CO2 and NaOH and influences pH in ways that can influence bed properties (somehow).
I find this quite fascinating and it actually has given me direction towards investigating an anomaly in our espresso program. We have 10 locations using largely the same equipment and the same coffee, however one of these locations has always demonstrated higher flow rates than the rest which are pretty much the same. This one location happens to be in a part of town with very old municipal water pipes... hmmm! Come to think of it this shop routinely runs at marginally higher brew temperatures than the rest (0.3F - 0.5F higher than other stores), which would make sense if the water has a slightly lower bicarbonate alkalinity (adjusting brew temp based on the inverse relationship between brew temp and alkalinity when alkalinity is already quite low... < 40 mg/L). Awesome... that would seem to explain this anomaly. Thanks Illy!
Update 09/12/09: This morning I did some preliminary testing of the above theory. At three of our locations pulling the same batch of roasted coffee on the same equipment with similar use & maintenance histories, the one shop that I noted demonstrates higher flow rates required a 0.7F to 1.0F higher brew temp setting over the other two to achieve the same taste balance. This is in accordance with the theory that the flow rate is due to lower bicarbonate content in the brew water.
I still have to get an actual measurement of bicarbonate alkalinity at this location relative to others as there may be variables I'm not considering here.