It is clearly a bad proposal.
Selling =/= dissenting
Your argument that it is a good proposal lies in the idea that selling indicates people (a) do not believe in Optimism, (b) do not value participating in governance.
This is reasonably obviously not true for all sellers. Other reasons someone might sell include:
- They think the valuation of OP is currently too high and they can wait until it becomes aligned with their expectations in order to participate in governance, and thus have a larger governance share.
- They think the valuation is okay, but they expect a post-airdrop dump. They want to buy lower and they think they can sell quickly before this “airdrop dump” happens and rebuy more OP.
- The airdrop is meaningfully large for their net worth, so selling some tokens makes a big difference to them, but they still believe in the future of Optimism.
- They sold on airdrop addresses and rebought on other addresses to anonymise their on-chain activity.
- They have tax obligations.
- They are selling 90% of their crypto portfolio because they are scared of a bear market, but they will rebuy in a couple of years.
- They are fucking rekt and this $ income helps them a lot, maybe one day if they are no longer rekt they can afford to participate in governance of the projects they believe in again.
Since it is possible to sell and still believe in Optimism and still want to participate in governance, permanently censoring addresses doesn’t seem like a good option.
Cost of user acquisition
Even if your argument were true, and all sellers did not believe in Optimism and did not value participating in governance, it is still not sufficient reason to censor their ability to participate in the future!
The airdrop spend equity (in the form of $OP) for user acquisition and user retention.
The focus on user acquisition and retention is clear through the airdrop criteria, where users and repeat users are heavily rewarded. They are even rewarded much more than DAO voters! 
Their tokens were earned for their Optimism usage. That was the criteria set by OP Labs (or whoever is in charge). Clearly, this was a reward for specifically using Optimism, something that OP Labs found to be an important metric. Plus, since users expect airdrops, users will speculatively use a product in order to gain the future airdrop. $OP spend is a customer acquisition cost for those users.
Now, imagine the future airdrop that you are censoring these users from.
Will the future airdrop target “usage” as a metric again? If a future airdrop were to target usage once again, that means OP Labs or The Foundation believe rewarding/paying for usage is still important. If you believe paying for usage (as bootstrapping mechanism) is important, how can you justify not including specific users? If you are paying people to use the product, it is expected that some of those people will sell by nature of the incentive!
If you don’t believe paying users is important in the future, then you should not define airdrop criteria to reward users.
If the future airdrop does not target “usage” as a metric, but instead targets governance-specific activities, and these addresses were eligible based on their behaviour from today until this theoretical airdrop, how can you justify not including governance participants because they once, historically, sold some tokens? They have done other actions that qualify them as good governors or participants. Is that overridden by historical financial activity?
If you want to airdrop only to people that will keep the coins, then you should not even make the token transferable! Of course, this would not be an effective customer acquisition tool, which is why nobody does that.
Why does the price matter?
The proposal is motivated entirely by the short-term price of $OP. The short-term token price action is detached from the performance of governance almost entirely.
The only people that care are price-speculators, traders and short-term investors. Governance certainly performs the same whether the price is $1.50 or $2. Sure, extensive price changes could make governances attacks on Optimism cheaper, but there is not a single mention of that in the initial proposal, and it is not really a concern given the size of the airdrop or the short-term nature of the sell-pressure from “instant dumpers”.
The valuation of Optimism will be defined by Optimism’s ability to gain traction, market share, being valuable to users, etc. Airdrop dumpers cannot materially impact that.
Does it matter if the price goes down for a day or two? Why does OP Labs care? It is the price you pay for using equity as a behaviour-incentivisation tool.
Is a ‘holder’ more valuable?
Perhaps you can argue that you can optimise the treasury and $OP equity spent by not including “potential dumpers”, and thus rewarding only holders or “long term believers”. You can argue that you’re giving free money to these evil dumping people for nothing.
But for this to be true, you would need to show that a holder-user more valuable to Optimism than a seller-user.
Is the number of long-term holders of OP an important metric to the success of Optimism?
Imagine two users:
-
User A is a regular, legitimate user of Optimism. They get all their friends to use Optimism. They have a lot of daily on-chain activity. They have no interest in participating in governance, so they sold their initial airdrop.
-
User B delegated their initial airdrop and didn’t participate at all in governance either. They still hold their tokens and their delegate votes every now and then. They have used Optimism once per month.
Which user is most valuable to Optimism? Which user should be rewarded in future airdrops?
I don’t see how the ‘holder’ criteria is necessarily the thing to optimise for, and don’t see how punishing historic sellers is the right mechanism for rewarding holders anyway.
Airdrop and token distribution criteria should match what it is incentivising. The previous airdrop, users were rewarded for being a user. If you want to reward holding, you can design mechanisms for that easily.
Ineffective measure
Even if you concluded that you wanted to do this, regardless of it being a bad idea, then the measure itself is ineffective. Users will simply change their address. It takes ten seconds. You do a lot of work, raise negative pr, etc for something that is completely ineffective.
If Optimism is to be the primary L2 of Ethereum, it will have millions of users. It will not be expected that every user is also a governor.
If you don’t want your airdrop to be sold, make it non-transferable. In fact, if you don’t believe your airdrop is a customer acquisition cost ($) then make it non-transferable and see what happens