There are already valid use cases for OP_RETURN, it only makes sense to fully support the feature. The only reason it's not supported now is because the Payments protocol came before OP_RETURN.

I give one example use case in the BIP. I agree that special wallet support would make the feature even better, but if someone tried to use Core the transaction would at least not be rejected. 

I've also been exploring this area with key.run (https://git.playgrub.com/toby/keyrun) and want the functionality for voting based on aggregate OP_RETURN value. *Not* to store data on the blockchain, but to associate content pointers with transactions.

I think that since OP_RETURN has already been approved and supported it doesn't make much sense for me to have to re-defend it from scratch here. 

On Mon, Jan 25, 2016 at 7:23 PM, Luke Dashjr <luke@dashjr.org> wrote:
On Tuesday, January 26, 2016 3:17:12 AM Toby Padilla wrote:
> I don't think every application of OP_RETURN could be classified as "spam".

Perhaps not, but in this context I cannot think of any non-spam use cases.
Use cases should come before changes to support them.

> I also don't think burning the value is going to dissuade anyone from going
> down that route. I don't think lost value is better for anyone.

Lost value is better because it has a cost to the spammer, and deflates the
rest of the bitcoins.

Luke