When the water in the cell walls evaporates the cells shrink. They shrink more at right angles to the grain than in the direction of the grain. If you exchange the water in the cells for something that doesn't evaporate but instead, cures, then your wood is much less likely to split or warp.
I have been studying on this problem and have read where some folks are soaking their wood in alcohol which mixes with water in any ratio (miscibility) but don't know how that alone would help. If after the alcohol soak you soaked the wood in an oil like tung oil or other wood worker's choice penetrating oil (miscibility of the alcohol and oil required) the end result would be replacing the water of the wood fibers/cells with a wood preserving oil. That should stop the shrinkage (and the attendant splitting and warping.) Acetone or other solvents miscible with both water and the epoxy (or oil or other preservative) could be used as well.
I'm also investigating means of getting a thin viscosity epoxy exchanged with the water in the wood. Probably something like the epoxy used in the West system of epoxy saturation which is normally used with dry wood, hence the need to find a multi-step (at least two) process to get the epoxy swapped out for the water in the wood before it hardens.
Pat