Quote

When will the testnet3 reach its max supply?

After all halvings. We are getting there, because currently, you have 4768 satoshis as the basic block reward, everything else comes from transaction fees. We already had 20 halvings, out of 32, so only 12 halvings left.

https://developer.bitcoin.org/reference/block_chain.html

Quote

All blocks with a block height less than 6,930,000 are entitled to receive a block subsidy of newly created bitcoin value, which also should be spent in the coinbase transaction.

So, we currently are around block number 4,300,000, and because of frequent blockstorms, we will probably get there soon. So, if April 2025 is around block 4,300,000, and November 2024 was on block 3,300,000, that means around million blocks in half a year. So, by roughly estimating things, I expect in 2026, maybe in 2027, we will get there. It mainly depends, if someone will accelerate the process by abusing blockstorms with timewarp, or not.

Quote

Bitcoin Testnet I think does not have a fixed maximum supply

It has almost identical rules to the mainnet, the only major difference is in difficulty adjustments, which caused blockstorms.

Quote

why is the supply limited?

Because the main goal of testnet is to test things. And if you want to test Bitcoin, and not something else, then you want to avoid changing rules, unless it is strictly necessary. And after many years, we can see, that allowing the difficulty to drop, was a mistake. Because that's why people created signet at all, if testnet3 would be more stable, then making signet may never be needed.

Quote

limited supply would increase that difficulty

Limited supply is needed to test, how the network behaves, when most coins are mined out of transaction fees. And by seeing, what happens in testnet3 in practice, we can see, that if you want to earn anything, then you are forced to confirm a lot of spam, because they are paying the biggest fees, and they are keeping testnet3 alive.

Quote

Each new version make the older ones unusable.

Not really. Each new version simply drops the support for the old network, but you can still use it. Nobody stops you from using testnet2 or testnet1 in practice, but there were not enough users, to keep any of these networks alive. However, in case of testnet3, it will probably live as an altcoin, at least for a while, because of quite significant chainwork, support of some centralized exchanges, and a long chain, maintained for years by many people, and also connections with some altcoins with non-zero value.

If anyone would really want to make the old testnet3 unusable, then it would require releasing for example testnet5, by mining on top of the same Genesis Block, as we have in testnet3, and reorging the whole testnet3 chain, by producing a stronger chain. Then, it could be really called "reset", and not "abandon".

Quote

and in most cases once your testing is done you are supposed to return the tBTC to the sources you got it from

This thing was never implemented, because testnet has no demurrage, and no coins have any expiration date, after which they could be burned, or taken by miners. A good question is: should testnet5 be created with such consensus rules?

Quote

At current pace and considering people sell/buy tBTC, i expect it'll happen within 10 years or less.

You underestimate, how many blocks were created quite recently. It would be the case, if we would have only regular blockstorms, and nothing else. But if you connect blockstorms with timewarp, then you can tweak timestamps, and produce a lot of blocks (to the point, that it will break some services, which use testnet3, as it was the case, when Jameson Lopp did it).

Current blockstorms in testnet3 are producing more blocks, than they were in the past, but it is still far from the maximum speed of six blocks per second, repeated during each difficulty adjustment.