Open source means "take your time doing it". More of a bless than a curse, hopefully. Wishful thinking.
It was always the case. If you are a software engineer, then you will handle Bitcoin easier, than if you are not. I can clearly see that in case of test networks, where by CPU mining I can get around 10% of all mined blocks, which is insane, if you note, that all of my blocks have the minimal difficulty, and there are many ASICs around. And it is even more surprising, when I can achieve similar results on both testnet3 and testnet4.
There's nothing to be "fixed".
One of many possible outcomes, is to move those transactions into test networks. I can happily confirm a lot of Ordinals on testnet3 or testnet4, if it will decrease the load on the mainnet, and will allow regular payments to be processed faster.
Due to the way the network is designed, even if you softfork and disable this "malicious" use, then Ordinals can appear again, in just another shape. For example, in UTXO.
This problem can be also fixed, but I am trying to postpone that kind of code, for as long, as I can. Fixing UTXO flood should happen after batching regular payments, not before them. Ordinals should fall into the trap of their current design, get themselves locked in their present flaws, and then, crushing that market will be easier, because then they will face the same problems as Bitcoin Core developers: you cannot change things, if they are grandfathered in.
A testnet blockchain is not an immune piece of data, which is guaranteed to be stored indefinitely.
The same is true for mainnet, but many people didn't realize it yet. And it is good to keep Ordinals' users in their blissful ignorance, for as long as possible, to then wake them up with a cold shower. If you want to know more, then read about "assume UTXO", and note, that Ordinals are completely unprepared for those kinds of things.
We already notice people migrating from testnet3 to testnet4.
1. It will be hard to kill testnet3. In the best case, it will be slowly dying altcoin. But I guess, that network will be maintained for a while, even if it will no longer be supported by Bitcoin Core, and even if existing seed nodes will be disabled. Some users will just not upgrade, and their local list of peers will be sufficient to keep it running, even without hard-coded seeders. And as usual, they will use addnode to connect to nodes, maintained by centralized exchanges. There were altcoins, which stopped working, and trading was not affected. This situation may also repeat there.
2. Getting the same outcome on testnet4 will take a lot of time. If you want to test things after many halvings, then testnet3 is your only choice, there is no alternative, because testnet4 started from 50 tBTC. Also, because testnet4 will have no blockstorms, then doing some tricks will be easier than on testnet3. Another thing is the new rule of 600 seconds, which will help CPU miners to mine more blocks than before, because it will push the time forward, and lower the network difficulty. I can also see that now, because I can mine more testnet4 blocks than testnet3 blocks, with the same CPU power.
It's crystal clear that the best selling point of Ordinals is that they're stored on Bitcoin, which will probably operate for a very long time.
Not that long, as you may think. They annoyed a lot of people. They abused the network to the point, where some big mining pools started censoring them. And they created way more UTXOs, than they should, so they forced some developers to deploy their code way faster, than they planned. They basically forced people to start spinning up new "utreexo nodes" to avoid being sued. And if they will abuse the chain further than today, then you will see block explorers, showing proofs, instead of sharing full transaction data, when their operators will start receiving letters from courts.