I think we may have to decide on a phrasing here. Being against Gavin can mean two things: "Against an increased maximum block size of 10 MB" or "Against any maximum block size". I think passing this 'problem' or rather decision onto the miners is the best way to go!
Right. I'm against his current plan of arbitrary increases. I'm not for NO maximum, but this is semantics -- there is also a max and min difficulty, etc. So, I'm for the max being de facto because of the protocol spec, as with difficulty.
Anti-spam measures need only exist to prevent DOS. (There might be an argument for preventing bandwidth-based centralization as well.) This can be managed via fees, which should be made algorithmic instead of arbitrary, as well.
To remind the thread,
most pools cap blocks at 750k right now. (If you don't know why they'd do this, you really shouldn't be commenting on blocksize at all, without more understanding of how this all works!) This isn't going to change with the 1MB cap removed. The economics of Bitcoin will not really change with a blocksize increase.
So, tl;dr my recommendation is
remove the 1MB cap; allow message length to be the current upper bound -- don't just drop another new arbitrary cap in place.