Supporting the long-view of Ethereum
Instead of creating tokens on an existing blockchain (e.g. Ethereum or any Turing-complete blockchain), some have taken the step of using their own blockchain (please see Venezuela’s “Petro” or others). By utilizing Ethereum, a token issuer has reasonable confidence that consensus algorithms have already been validated. By using a speciality blockchain, unless properly defended, you can expect there to be attack vectors (including the 51% PoW attack) to basically destroy the entire chain.
One of the most important aspect to blockchain technology actually has nothing to do with ledgers or token values, rather its most important aspect in my view is the nature in which it uses economic force to encourage / discourage specific behavior. For example, if a miner has sufficient hash-power to reach 30% or even 40% of the collective hash-power, rather than act nefariously, a miner will be motivated through economics to just mine and receive their tokens.
We really haven’t seen mass attacks on small hash-powered chains. Yet… But for sure it is coming. Like an incoming tide about to hit a sand-castle, it is coming. Count on it.
As a result, more and more tokens will be pushed to large hash-secured blockchains (like ethereum). At present, there are 125,511 ERC-20 tokens on Ethereum. Expect that number to continue to grow (for ERC-20 and its replacement ERC).