In the end I said "other" because I'd like to see a first-rounder or young talent (please, not Noah) come back for the presumably valuable 2010-ender in Brad.
Brad for Wallace is intriguing, but it doesn't look like Charlotte would be down for it. There are a couple of reasons, hashed out here before by many, that I would be a bit reluctant to get Wallace on a long-term deal - his tendency to injury, his concussion record, and being the eternal optimist, I want to see us develop Donte Greene into an all-star level inside-outside SF.
On the surface I don't really love Brad/Kenny for Marion, but there's a lot of cash to be saved there. All of Brad and all of Kenny for next year is $21M, and Marion comes out a bit lower this year, so we might save up to another $1M prorated in '09. So it is a bit tempting to let all that money go. I don't know what we do with that money, though, outside of make a run at Boozer.
The downside of trying to deal Brad/Kenny as enders for draft picks is that the teams trading for them will want to send bad salary back to us. And that's $21M of bad salary for at least the '10 and '11 season. And if you ask me would I rather have:
A) Two middling picks (because they'd likely be lottery protected) and maybe two mildly promising kids, and some bad salary
B) Miller and KT to the end of their contract
C) A partial year of Marion
Then A) seems the obvious choice. But let's look at approximate marginal cost, relative to the cheapest option:
A) At least $40M and reasonably $50M or more (not including the salaries of the draft picks) spread out through '10, '11, and possibly later
B) About $21M for Brad and KT in '10 (about another $1M for Brad/KT in '09)
C) $0
Then it's just not so obvious. You're paying out a marginal $10M+ per kid/pick, and if you play the Pritchard game properly, picks can be bought for a lot less.
Man, this makes me lean towards Marion and just getting the cash off the books...