I'd be careful. The soil surveys are more accurate than they are generally given credit for.
Editing this to clarify that I think the error in the soil maps has more to do with the location of the boundaries between soil types (due to the number of samples taken)as opposed to the type of soil, what it is composed of, and what limitations it has. I'm not a nuttygambler, but I'm sure you can make it work by applying the appropriate planning and a sufficient number of $.
