Is there always a security tradeoff with Plasma sidechains?

Was reading a bit about progress on plasma today – Loom and Ignis specifically. I’m trying to understand the security tradeoffs with plasma chains, and have a few open questions:

1. The plasma chain specs I’ve seen so far all use DPoS for consensus of the Plasma chain itself. Loom describes Dappchains as “EOS on top of Ethereum”. Regardless of if there is a secure transfer gateway or not to Ethereum mainnet, doesn’t this architecture just inherit all of the issues with DPoS? I can see how this tradeoff would be fine with social and gaming use cases, but wanted to make sure the transfer gateways to mainnet don’t offer any sort of increased security in some way.
2. Ignis ([****](**)** are using SNARKs to provide proofs of computational integrity to the main chain. If a Plasma chain is DPoS, is there any way these SNARKs could be forged by malicious validators? I’m assuming there could be censorship among transactions internal to the plasma chain, but I’m wondering if there can be any sort of forgery with the link to the main chain.
3. Is there always a securty tradeoff with L2 plasma chains. Are there any architectures that would allow a high security Dapp to run on a plasma chain?

Feel free to point me to articles, etc. if there is an obvious answer to these questions. Thanks!


EDIT: To clarify, I’m mainly thinking about transactions that interact with smart contracts and involve some sort of state persistence or computations – not so much about just transfer of value on the plasma chain

View Reddit by offthewall1066View Source