Walters. Weeds: Control without Poison, has some interesting ideas. It is part of the Albrecht, Acres USA diaspora.
There are some good basic analysis about soil saturation in the wetland delineation literature.. the USDA PLANTS database has a regional wetland indicator code that is used to delineate wetlands. That is relatively data driven and based on lots of observations, and it shifts based on region.
There are indiciator species in the Veg Ecology literature... Klinka and Krajina out of BC for example has a Indicator Plants of Coastal British Columbia... http://www.ubcpress.com/search/title_book.asp?BookID=1443
nice and expensive... They only differentiated among four axes... and don't get into plant presence of vigor predicting soil composition.
The problem is that what a plant indicates differs by climate. A plant that might indicate dry conditions in the PNW might indicate wet conditions in California. Climate and resulting pH affects nutrient patterns. I have looked for the basis for all the claims of plant indicators... I've seen some of those lists, and I can't help but wonder... how do they know this? What is this based on? My faith is generally low when someone tells me of a clear cut species-soil condition correlation... there are just too many factors affecting what grows--I've seen veg structure shift so much over time through succession or based on 'founder effects' (who got there first).