I really like the piece balance in this variant.  No one type of piece is
very strong (I think the rook is strongest), but there are a number of
different pieces with different abilities, and I think they are balanced
quite well.  For instance, I disagree with the variant notes about
liberating the giraffes; I believe they are limited the way that they are
to prevent them from being too powerful; otherwise, setting two of them
side-by-side could effectively control four whole files, which is just
too
much power for two pieces to have, even on a larger board like this.  As
the rules are given, though, they are not so overwhelmingly powerful
because they can be blocked easily by relatively close pieces, and so to
retain their power have to keep some distance from the action, or their
close range has to be protected by other pieces (a job I suppose rooks
would be well suited for, if they are not busy elsewhere).

I thought at first that the war carts and elephants would be weak, but
upon further inspection they are useful and complement the other pieces
nicely.  
My only quibble with the elephants is that their moves are not very
elephant-like (unlike the giraffes, whose move makes sense for that
animal
in a couple of ways).  But that is a very small thing, and the elephants
are as in Shantranj in any case.

The general and the vizir seemed weak to me at first, and then I realized
that without castling, the king will need something to hide behind, and
that is why those pieces are not very mobile.

Regarding the guard variant, I'm not sure whether it's necessary,
though
it does seem like it would help.  The elephants and camels and war carts
are balanced to some extent by the pickets (although I do like the idea
of
playing the pickets as regular biships, partly because it helps fortify
their squares, and partly for symetry, since the horses and rooks are as
in modern chess), and in any case both players are in the same bind for
the squares those pieces can't cover.  Then again, the way pawns promote
could upset the balance further; for instance, if a player were to get
his
pawns of camels, war carts, and elephants all promoted, he could have a
strong position on those squares then.  On the other hand, getting one or
two of those promoted on the opposite squares, plus the pawn of pickets,
could have the reverse effect.  I guess I'd have to play it a good bit
both ways to be sure whether adding a middle column is a worthwhile
improvement or not.  I'm thinking it might be.

I believe I would prefer to allow the pawns the initial-double-move
option
as in modern chess, mainly to help get things going at the beginning,
since
the board is a bit large.

I think if I wanted to simplify the rules a bit, the first thing I'd do
is toss out the citadels and the special draw and stalemate rules, and
use
regular chess rules for those things; I don't think that would upset the
balance of the variant at all.  I might also simplify the pawn-of-pawns
promotion rule also, perhaps allowing it to promote (the first time) to
any one formerly-captured friendly piece the player wants back.  But
these
are minor quibbles.

On the whole, this seems like a very nice, balanced, playable variant,
with some fun pieces.