Ones I missed, but are actually worth mentioning when I think about it:
Pawn-bishop: Probably more of a hindrance than a help in the opening and early middlegame; but you might want to save some to the endgame to attempt to promote them. Also, it's low value means you can use it for blocking tactics you wouldn't attempt with a regular pawn.
Pawn-rook: Should probably be thought of more as part of the architecture of the board than something a player controls.
Queen-pawn: Not too much weaker than a standard queen, but imo still worth 8 points or less.
Rook-pawn and Bishop-pawn: Practically the same as the standard versions of the piece.
So usually when making up new chess pieces, varients tend to combine existing pieces, such as rook+knight.
I was wondering what might happen if you subtract the movement of pieces, instead.
I'm interested in everyone's thoughts, but I'll give my own too.
Queen-King: I actually think this would substantially weaken the queen, since it loses most of its checkmating potential. Also interestingly it might be a piece that's stronger on the edge. Perhaps a bit weaker than a rook?
Rook-King: Would make rooks more awkward to use in endgames. About as strong as a knight?
Bishop-King: 1.5-2 points maybe?
Nonroyal king-rook: aka a ferz
Nonroyal king-bishop: aka a wazir
Nonroyal king-pawn: Would be stronger on the far side of the board. Worth 1.5-2 points?