According to this Wikipedia article:
Wood warping costs the wood industry in the U.S. millions of dollars per year. Straight wood boards that leave a cutting facility sometimes arrive at the store yard warped. This little-understood process is finally being looked at in a serious way. Although wood warping has been studied for years, the warping control model for manufacturing composite wood hasn't been updated for about 40 years.
Factors contributing to the warping of wood include plant species, temperature and temperature changes, humidity and humidity changes and wet-dry cycles.
I'm not interested in which species are more prone to warping, how to prevent warping by treatments, etc. Nor am I interested in the biological processes of living or freshly cut wood. I'm interested in the processes happening over time on a molecular level that lead to macro-scale warping of wood. Water, in particular, seems to play a major role in the warping of wood, implying that possibly the lignin and cellulose matrix of wood is disturbed by the formation of hydrogen bonds. But if this is the case, or whatever the case, why would the microscale effects not cancel out rather than causing macro scale warping in a particular direction? This may be more of a physical than a chemical question, however.
What are the chemical process responsible for the warping of processed (as opposed to living or freshly cut) wood?
A link to an interesting and relevant video was given in a comment, in which several types of wood were placed in a chamber which was evacuated of air, then pressurized with ammonia gas. The video also gave a link to a patent for this process. The following is an excerpt from the patent:
The rigidity of wood is the result of crosslinking between adjacent cellulose chains by water molecules which hydrogen bond between sites on adjacent cellulose chains. Anhydrous ammonia is extremely reactive with water, and it is believed that anhydrous ammonia scavengers water from wood, breaking the crosslinking between cellulose chains, and permitting the cellulose chains to slide, which renders the wood very flexible so that it can readily be bent or twisted. When the ammonia volatilizes from the treated wood, water vapor reestablishes crosslinking and rigidity.
The reason for this edit is to suggest that the disruption of hydrogen bonding between adjacent cellulose chains, when the wood is dried then rehydrated, for example, may also play a role in "natural" warping. I'm going to look into this more and I hope it gives somebody else a clue to a good answer.