P-G: I've been thinking about your recent posts. They are worthy of discussion, whether or not merit. Indeed, many adaptations could be made to the grandchess set that conceive "grandchess variants" just as heretofore we have witnessed hundreds of registered chess variants.

Of course, there is a sizable difference in these fields. Chess is centuries old and, especially in the modern era, has been deeply analyzed and become so overmechanized that the highest level games are almost automatically drawn. (This perceived exhaustion of chess was the driving impetus to create variants of it!) Grandchess though (while itself known as a classic chess variant) is just 25 years old, and has undergone no such methodic exploratory transformation. It probably never will, and certainly won't in our lifetimes, so the necessity to develop variants thereof is much less pronounced, although the ideas can be of interest.

The cannon is a good tactical game piece, but its lack of popularity outside the Orient may be due to its obtuse nature in comparison to real battlefield situations. The diagram array you presented appears to be more like an entirely different game construct, rather than an offshoot of grand. You combine elements of xiang-qi, shogi, and grand, but you also placed the Kings on the back rows which (unlike grand) is more akin to Janus, embassy, Capablanca's, and perhaps other ideas. You also expressed a desire to deviate from pieces based on Knight movement, or at least some of them. I don't know what to attribute this to other than personal taste.

Some thoughts on the promotions:
-- First, you suggest that a pawn be promoted to a gold general. I'm not a big fan of shogi, but I like the gold and silver generals; the gold is slightly stronger of course. But, on the 10 squared board among the grand array, they would be sharply ineffective. It is hard to imagine how they could even outplay the Knight, which is unquestionably the weakest piece.
-- Second, with this suggestion you directly contradict the grand standard of promoting only to an already captured piece. Some leeway is allowale here, since variants are mostly about rule and/or piece changes, but it makes your idea that much further removed from grand.
-- Third, you state that a captured gold would revert to a pawn, but this is superfluous since there are no piece drops in grand, unless you want to play it like chessgi (here called "loopchess") but that changes all the game dynamics and unearths a Pandora's Box of whether to allow pawns to be dropped on the eighth or ninth ranks. Or, perhaps you were merely stating a condition of shogi, I couldn't tell.
-- Fourth, Walter says he would do away with eighth rank promotions. You wisely refrained from embracing this notion. Shortly after introducing his game, Freeling explained that he had tested limiting promotions to the ninth and tenth ranks, but decided it better to include the eighth rank as well. Even at that, the suggestion to abolish eighth rank promotion has been made before, and been met with disapproval from most grand players. It may only take four moves to promote a pawn, but that doesn't make it easy.
-- Fifth, By the time one side is able to achieve a promotion, they usually have a winning edge anyway, so a rule change based just on promotion would have scant impact. Of course, if that were incorporated into other changes, as you have outlined, this would be a different case.

Ultimately, I suspect that variations of grand (if any are to become popular) will be those that (like chess variants) make one simple rule change (instead of an eclectic sort) that is easily understandable yet still provides prolific and dynamic play. But, since the mysteries of grand itself are still largely untapped, it seems likely that this site (or others) will be reluctant to add modifications to the basic game. So, it might be that the best we can hope for is to have fun with these new ideas by tinkering with them on our own quarters, or playing against friends.