If the issue is spam control, then wouldn't that be negated a variable fee based upon based upon the level of risk that spam may be contained within the transaction? Such a variable fee in this case is non-linear to the amount of 'risky' data attached. It would remove the arbitrary nature of 40 or 80 bytes.

It is difficult to predict what is necessary or unnecessary in the future. What if I wish to publish into the ledger the the daily fixings of the history of a 500 million dollar notional interest rate swap. Perhaps at that juncture in the future we would decide it was inappropriate to be stored in the blockchain. However, if it was appropriate I would certainly be willing to pay a larger premium to place it into the ledger.

I never said I didn't want the Bitcoin protocol to change.
On the contrary, I support extending it to do what Counterparty wants.
But such extensions are slow-moving right now, and take time to implement properly.
I also understand Counterparty wants a solution "today".
I agree the 80-byte OP_RETURN is a good short-term way to do this.
Deploying a whitelist to miners, to accept these Counterparty transactions can be done within a few weeks.
But deploying a default relay policy change requires months (releasing a new version of Bitcoin Core, and the slowest part: waiting for all the users to upgrade).
Thankfully, there is an immediately available workaround to not having the default relay policy "on your side":
Just have Counterparty participants relay their transactions to nodes running the updated relay policy.

So, recommended course of action:
Immediate-term:
1. Write Bitcoin Core patch to whitelist 80-byte OP_RETURN-based Counterparty transactions.
2. Deploy patch to major mining pools, and open merge request with mainline Bitcoin Core.
3. Begin using OP_RETURN Counterparty immediately; use addnode to get it relayed to miners.
Short-term:
4. After discussion, patch is merged to Bitcoin Core.
5. Bitcoin Core 0.10 is released with a default relay policy accepting Counterparty transactions, and addnode is no longer needed.
Long-term:
6. Counterparty developers discuss future plans with Freimarkets developers and others interested in this kind of functionality.
7. Interested developers figure out the best way to do everything, probably including using merged-mining, side-chains, and other things that are impractical today.
8. Interested developers implement new system, and write a BIP documenting it.
9. BIP gets reviewed.
10. Counterparty users upgrade to new version based on BIP.
11. Everyone gets a break.

Hopefully that clarifies my position.

My understanding is that Counterparty is functioning, right now, using Bitcoin as a transport layer.  In order to do so, it must be using existing, accepted features of Bitcoin.