It's not fair that other buyers will profit and he wouldn't.
At the same time, it wouldn't be fair to other Agora holders to let the issue unresolved.
Here are the constraints I can see:
1. Dor should keep at least as much profit as other buyers who didn't step up
2. Other Agora holders should be made whole for the accidental dilution of their stake
3. No funding money / sales proceeds should be taken from Tau/Agora project to resolve the matter as this would be counter-productive to everyone.
4. Ideally, Dor should get some reward for his transparency and willingness to arrange the situation,
Here are a few ways we can do that without much drama:
Proposal 1: prorated holding increaseWith the help of Dor (log of his trades) and the trade history on Bittrex, we can calculate the average price (let's call it P) of all the trades of other buyers (other than Dor) and the total quantity of these trades (let's call it Q). We don't know how many actual buyers these trades correspond to, so we have to assume that all other trades belong to a single buyer.
(1) can be satisfied by allowing Dor to buy Q tokens at price P per token (minus what he already paid) and return the rest of the tokens to Ohad. If there is a single other buyer, they are even. If there are more (which is very likely the case) Dor gets the best deal of everybody in terms of total notional profit.
(2) following (1) the total dilution is 2*Q. Say the total amount of Tau already sold to investors under the crowdsale is T. The new money mass following accidental dilution is T+2Q. The increase rate of the accidental dilution was therefore R=(T+2Q)/T. Holders can be made whole by multiplying their holdings by R when you issue the new token. It may seem that increasing every buyer's holding equally is the same as doing nothing as it doesn't change the relative holding of every buyer with respect to the current float. But since there is a fixed cap of 42M to the max supply, and later buyers won't have this scaling applied, this still helps compensating earlier buyers in the longer term.
(3) Since only tokens (that would have been destroyed at the end of the crowdsale anyway) are spent by Ohad from the crowdsale fund, this doesn't affect the funding of the project.
(4) IDNI / Agora foundation's own pool of tokens can be used to pay a discretionary reward to Dor for helping. Since these funds would have been issued anyway at the launch of Agora, this doesn't affect the money supply long term for crowdsale buyers.
Proposal 2: prorated saleAnother approach would be to give to everybody the same opportunity to buy tokens at a lower price prorata of their current holdings so that what started as an outage endup being a firesale that will benefit everyone including the project's finances.
We calculate the average price of the transaction. If Dor is keen to cooperate and buy his tokens, P is the average price at which other buyers got their tokens. If Dor is less or not at all keen to cooperate, P is the average price at which every buyer including Dor got their tokens. Now, to calculate cap on how much each holder can buy, we take R = Dor's Recent Purchased Quantity / Dor's total holdings (I understand that Dor holds a lot already, so this ratio shouldn't be that high). If T is the total current supply of tokens sold in the crowdsale, you start a sale for maximum R * T tokens. Each agora holder can buy up to R * balance of his address at price P. You will need to make a snapshot at a specific date and use balances in the snapshot. Holders must move their funds to a Omni wallet before the snapshot. Additional tokens are to be sent at the same address as the one used to calculate the cap to avoid cheating. You should put a time limit on the offer, long enough for people who care to have the time to take notice.