The number of coins left by the SN would be fairly divided by the number of addresses already in use that hold the coins.
Even if coins would sit on "1000000 OP_CHECKLOCKTIMEVERIFY OP_DROP OP_TRUE", and someone would want to distribute them, then still: sending the same amount of coins to all existing addresses is very far from being "fair". For example because one person can have more than one address. And then, someone who owns 100 different addresses will receive less coins, than someone else, who owns 1000 different addresses. Assuming that one human owns one address is far from reality, unless you want to globally push everyone through KYC.
Is it possible or not?
Possible? Technically yes, but practically, a lot of people will be against it, so it will be just some altcoin, which people will happily sell for the old BTCs instead.
You're all screwed if Satoshi goes down and sells his coins.
You always need two parties to make a transaction: if there is a seller, then there is also a buyer. Even if Satoshi would start selling coins here and now, and would want to sell everything immediately, then still: some people will just get cheaper coins, and after that, everything will just go back to normal. Because after happily buying them cheaply, they will have no reason to sell them low, especially if they would know, that they were owned by Satoshi.
There were many altcoins, where creators did "pump and dump". And guess what: if the project is useful, then after being dumped, it will recover. It is just a matter of time.
How confident you are that those 22000 address belong to Satoshi rather than someone else?
It is only a pure speculation. For example:
http://satoshiblocks.info/There is block number 127, which has unspent coins. How do people know, if it is mined by Patoshi or not? They simply guess. It has extra nonce equal to 6, and the next block uses 9. On this chart, you can see, that block 127 is marked as Other, and block 128 is marked as Patoshi. Why? Because the creator of this chart guessed so. But there is no proof, that it is the case. Both coins are unspent, so there are no traces.
Some people wanted to really re-assign some of these coins, and guess what: they didn't, because there is no single address with one million BTCs from 2009. There are many separate ones, and nobody can be sure, which address is owned by whom. And even if we assume, that Patoshi pattern really groups the same addresses together, then still: each time the counter goes to zero, there could be a different miner.
What counts as participant? node ? hashrate ? something else ?
There are a lot of ways to split the coins. Even if there would be something, spendable by anyone, then still: splitting it honestly is a hard task. And reaching consensus on that would be difficult, even if coins would be locked on OP_TRUE with some timelock.
