On the scalability issues of onboarding millions of LN mobile clients

In a discussion on the bitcoin-dev mailing list, Antoine Riard suggested that the trust-minimization of Bitcoin's security model may be shifted by Lightning Network (LN) where fast, affordable, confidential, censorship-resistant payment services may attract a lot of adoption without users running a full-node. However, Luke Dashjr strongly disagreed and stated that this cannot be shifted as it would compromise Bitcoin's security which depends on the assumption that a supermajority of the economy is verifying their incoming transactions using their own full node. LL, another participant in the discussion, questioned this claim and asked how not verifying incoming transactions properly could affect anyone else. He argued that getting scammed through fiat currency transactions did not negatively affect the currency overall. The links provided by Dashjr referred to "miner control" as the culprit, where a light client could follow a dishonest majority of the hash power chain. LL wondered where the side effect onto others on the network came from if he were to receive a payment in some miner scamcoin on their dishonest fork (but he thinks it's BTC because he's running a light client).