I've been molding with water and ammonia since 1974. Don't remember where I read it but the article I fot the info from indicated the ammonia softened the lignin in the wood (the pitch in the grain for want of a better description) allowing a greater degree of molding without breakage. The most important aspect of the ammonia was supposed to be the fact that the lignin rehardens after evaporation and helps the balsa to lose its *memory* and remain in the new shape.
This appeared to be true. My 1977 concours winning Citation had all the top and bottom blocks *and* the bubble canopy molded in this fashion. The canopy was particularly impressive in that it had no tendency to spring back to its flat shape and, to this day, it sits up in the attic loooking like it did when first finished. It has, to the best of my recollection, no internal structure to support it.
Re strength. The current Final Edition was built with the first ever (I think) molded leading edges using the water/ammonia mixture. This is the same plane that was the Great Expectations which flew in the 1994 World Champs and was completed the previous year. I don't know how many flights it has on those leading edges but it's a bunch. To date there is no indication of failure. This is true even though the ship was stripped to bare wood and then had the wing leading edge and center section sheeting covered with CF veil while refinishing in the purple pond scum theme, thus making a potentially very hazardous stress riser right at the fuselage. Again, no apparent damage has occurred.
(Please note re the above. This ship was built with three carbon fibre reinforced spars and is very rigid in spanwise flex as a result. I do *not* advocate doing what I did on a wing which doesn't have such substantial sparring. A failure at the stress riser of the fuselage is predictable for such a condition.)
If, in fact, strength after molding isn't an issue for ammonia based molding I think one could say the following. The use of ammonia does allow a greater degree of curvature before splitting (I do top and bottom blocks all the way down to the tail post using either multiple layers of 1/32 or a very good A-grain 3/32 single sheet. The resulting radius at the tail post is about 1/8 inch or a touch less.
In addition, I believe the ability of the part to retain its shape is enhanced. When the Ace bandages are removed from parts molded with the ammonia mixture the part has to literally be "popped" off the mold due to its willingness to remain in its new shape.
I'd say if your project requires serious radius curves or any sort of compound curving the ammonia probably makes sense. For more modest shapes which will, in the final analysis, be supported by bulkheads or ribs, etc. the ammonia is probably overkill.