If something is sold, there are two parties: the seller, and the buyer. If premine creator will sell the coins successfully, then it would mean, that someone will want to buy them. Which means, that premine will just land in different hands, but that will be the only difference.

If you want to see, what can happen with premine, then you can trace the history of some altcoins, which did it.

Also, the client already has some centralization aspects, for example in signet, where you can pick a different signet challenge, and have your own test network. In the same way, something like "premine challenge" can be made, and then, resetting the official test network will require only tweaking a single parameter, and test networks can be as easy to launch, as making new signets currently is.

If the only way to make test coins more worthless than they are, would require releasing new versions more frequently, or changing network conditions on the fly, then developers will go in that direction, because why not? If people are stupid enough to trade coins, which are designed to be worthless, then by making these coins really worthless, all "investors" will be harmed, when they will see, that they bought a lot of coins, and nobody else wants to keep using them, because it is "just old testnet12345 coins, and everyone switched to testnet12346".

For now, when I want to test things, I use some of my own code, where everything is resetted every 10 minutes (or to be precise: every time, when there is a new mainnet block).

If a dummy test network, made by Bitcoin developers just to test things, is worth more than serious altcoins, with serious developers, who want to make their coins valuable, and promote them everywhere, then it sounds very bullish for BTC, and in that case, altcoins won't have much chances to replace BTC in the future.

But it also means, that existing rules, like minimal difficulty rule, are not enough to make things worthless, and new testnets should be much more buggy, to bring their BTC value much closer to zero, than it is today. Which also means, that if premine will not crush the price, then you can expect more naughty changes, with the purpose of making test coins more unusable, than they currently are.

My current setup simply traces all mainnet headers, and tries to use Merged Mining, to make a new block on top of that. If I hit something (which didn't happen yet), I would get 3.125 BTC plus fees. But most of the time, I can get just a new test block, and use it for my own purposes, because I know, that it is worthless enough to be used.

Testnet was literally created to be raped, because then, you can push it to its limits, see in practice, what happens, when you break some things, and then implement protections for the main network.

Of course, testnets were made by aliens, and people just found them, and simply joined.

Seriously: it is quite obvious, who created each testnet, and why it was made. You can read posts, made by testnet creators, and follow their way of thinking. They definitely had some goals. And note that when testnet1 or testnet2 was created, there was no regtest, no signet, and people just experimented in different ways, to test some things.

So, the currency by itself can have no goals, in the same way as dollar by itself has no goals. But someone created that dollar in the first place, and it is quite obvious, that people can have goals. And the same is true here: there was a need to build some testing environment, where you can have some nodes, willing to participate in a network, where you can test edge cases, without taking real funds from real people. That's the reason, why test coins should be worthless. If they gain some value, then you create problems, because you can no longer test "what happens, if I do this naughty thing". Which means, that you can no longer continue testing, and you have to switch to something else (which is why test networks are resetted).