As noted, that means mining would end in 4 months assuming the initial target of 1B total coins remained intact. So after 4 months we'd be left with video producers and viewers being rewarded with just the tail emission. I think that would kill the project quickly. They are, after all, the people who will decide the fate of the project in terms of widespread adoption of the IPBC platform, not miners or investors.
It seems like you haven't looked into IOTA or how many IOTA tokens actually exist. Simply moving the decimal place will easily move the total coin supply back to 1 billion total.
None of this even begins to address how you would accomplish paying the miners between the fix and start of your proposed hardfork when working with an anonymous blockchain.
I already suggested paying an additional 9600 coins to the receiving address for each coinbase transaction that pays only 400 coins. If pools were already in place they should have records of block payouts per coinbase. If not, worst case scenario is you can't re-distribute those 9600 coins per block (however many blocks were affected by this bug), and those miners are at a slight disadvantage because they mined 400 coins per block for a few days. That's a lot better than having a coin forever known to have an instamine "mishap".
The other solutions are worse because they have long-lasting drawbacks. Fixing the problem by buying out the coins only gives an incentive for million-coin holders to not sell because they know another whale is buying up coins and promising not to sell. The only winners there are the early insta-miners who hold instead of selling to the buying whale. Leaving the mining reward low at 400 coins per block is also pretty bad considering only 100m coins are mined per year (before a halving), although, I'm not sure what the halving schedule of this coin is. If the halving schedule happens quickly, the instamined coins are a bigger problem, because it will take longer than 2 years to reduce the instamine supply below 10%.