"think it's actually a product of water-logged soil. Regardless of the actual function of the knees (seems to be a debated subject), " All the experts seem to agree that correctly watered trees do not produce knees." (Bald Cypress Knees) So it seems that Bald Cypress that get "too much water" produce knees whereas those that don't do not (" trees grown on drier sites may lack this feature." see Wikipedia's Taxidium distichum)"
False according to research.
Baldcypress: The tree unique, the wood eternal is a must-read if you're into BC. The book, whose author was with the University of Louisiana's botany department, did research in the swamps on the tree, accumulating a vast amount of information, research and observation. The author says knees are observed in trees on dry land, as well as in more submerged locations. He suggests that many landscape specimens haven't been in their locations long enough to produce substantial knees. He also notes that big trees growing in deep water lack knees.
He also says there are two kinds of knees--those that are outgrowths of roots that represent typical upright knees and a type that results from downward growing root that turns upward and once above the soil, turns back immediately downward (which would approximate the "bent root" artificial knees created by bonsaists)