Indeed a very complex question to answer. One of the most interesting parts of the paper is the discussion of the various types of scam — simple rug pull, sell

Indeed a very complex question to answer. One of the most interesting parts of the paper is the discussion of the various types of scam — simple rug pull, sell rug pull, smart contract trap door of various flavors. From their paper, I think sympathetic to your POV: > For example, it is not clear that cryptocurrencies such as Doge or Shiba have any use case or intrinsic value, but they are among the most popular meme-coins. In our framework, we say that a token has no intrinsic value or use case if the developer knows that the trading price with respect to USD will eventually be zero. In other words, a tradable malicious token in Uniswap induces a zero-sum game between the users and the developers, i.e. the incentives for the investors are not aligned with those of the token creators. Therefore, the main difference between malicious and non-malicious tokens is the developer’s intentionality towards the token. One of the main problems of these definitions is that it is unfeasible to distinguish between scam tokens and under-performing or abandoned projects without accurate off-chain data.