Why renaming the Kirin and Phoenix?

I never liked the name 'Kirin' very much. (It is meaningless to westerners.) And 'Phoenix' collided with the naming theme, as I wanted to reserve bird names for the pieces with 'flying capture'. The flying Q and R needed to be separated by some low-valued obstacle to prevent they would attack each other in a symmetric initial setup. And I did not like to stack them behind the flying B, as that would leave the suffocated King and Terrors exposed to frontal flying attack. Ground-based units that are able to interfere with overflying birds must use projectile weapons, which led me to Archer and Spearman. (I would have liked Crossbow instead of the latter, but that was not in the alfaerie set.) One excuse for renaming is that in this game they are not 'normal' Phoenix and Kirin, but passively guard the skies as well, unlike any other non-flying piece.

Also the board is very dense, more than the 50% I'm used to work with. I don't know if this is a factor favorizing shorter games.

I guess this is a Shogi influence that trickeled into the design, as there filling can be near 67%. How this affects game length is a subtle issue. More pieces tends to increase the length, larger (and especially deeper) board would too. With 50% density the 80 pieces that I have would require 18x18 rather than the 16x16 I used, but switching to 18x18 would leed to longer games. Decreasing the number of pieces to 64 would shorten the games. But this would be at odds with the goal for making a huge variant. I guess it depends on what you use as the metric for the size of chess variant: number of pieces or board size. For a given number of pieces high density shortens the typical game by more intense combat.

As I mentioned, it is mainly the depth of the board that counts. Or even more accurately, the initial distance between the armies. I dislike distances of more than 6 ranks; this drags out the opening phase, just to transport the army to where they can engage the opponent. For ~80 pieces you would need to have a 27x12 board to get 25% fill (each army 3 filled ranks, and 6 ranks in between). This seems too extreme. (Although it would solve the problem of Pawn scarcity.)

An alternative would be to not completely pack the armies in the initial setup, but starting them in a more advanced position in a more vacuous setup. E.g. use 18 x 18, but spread the 80 pieces over the 108 squares of the first 6 ranks for a 49.4% overall filling. That would leave ample free space within the camps to quickly develop pieces that started in the rear. So perhaps this is the superior solution. It would make the appearance quite different from the giga-tera-exo series. But I like the idea!

I'm not fully convinced that it is necessary to forbid the Warrior to capture normally. Warrior is maybe not a good name if it doesn't fight much.

Good point. This was an (admittedly ugly) solution to the problem that a Pawn protected by another Pawn cannot be eliminated without incurring a loss when you can only attack it with pieces that are worth more than 2 Pawns, no matter how many attackers you bring to bear. The only viable way to threaten it is with another Pawn. Which takes a long time when these Pawns have to cross a large board to get there.

Now the Terror can pick off protected Pawns with its Advancer capture mode. But against a closed rank of Pawns Advancer capture does not end on a safe square. So if a player can support the foremost Pawns on their original rank by moving up another wall of (diagonally capturing) pawns directly behind it, it would either take large sacrifices to break through it even with the help of Terrors, or very many moves to organize a Pawn storm.

Of course this is a self-inflicted problem, due to having multiple pawn-like pieces in the same file. But I wanted to include some 'spare' promotion ability, and the pieces providing that would have to be placed somewhere. One way to solve it would be to make the Warriors more powerful, and therefore more valuable, so that Warrior + Pawn is worth more than a typical minor, and capturing a Pawn protected by a Warrior with an abundance of attackers no longer is a sacrifice. But I don't want this extra power to result in the ability to find an easy path to promotion. This could be done by giving it extra captures, but I also don't want such low-valued pieces to be able to mutually protect each other, as this would allow them to form hard-to-break defensive structures all by themselves. Pieces with many captures and few non-captures tend to favor defense over attack.

An alternative would be to provide sufficiently many pawn-valued pieces that can capture into a tightly knit double wall of pawns with disregard for its own survival. And that do not take a long time to cross the board. My original idea was that the Grasshoppers would satisfy this role. But it seems they are too strong, and too easily find prey amongst the opponent's more valuable pieces. The requirement of fast movement and low value is somewhat contradictory, but Grasshopping does provide it. True Grasshoppers have much to dangerous captures, though. So perhaps a divergent piece that moves as Grasshopper, but captures as a (forward) Alfil (or even Dababba) would do it ('Ram'?).

Adding extra pieces to specifically destroy pawns seems to defeat the purpose of adding extra promotable pieces, though. Unless there are fewer of those. I suppose this is justifiable, because the attacker can choose its point of attack, while a defender would have to keep up the defense everywhere. So 16 Pawns, 8 Warriors and 4 Rams might be a good mix.

If the Warriors become more or less normal Pawns, they could be started in the wings of the 2nd rank, with King & Rooks on the 1st rank, so that castling would produce a normal King fortress after the pieces starting in front of it have left to engage in battle.