It's a continuous usability vs security tradeoff spectrum.

Spreadsheet: just trust a third party

Ecash mint: same but party can't track/spy

Element-like chain with CT: same but party can't inflate/fractional-reserve either

Statechain: same but party can't rugpull without colluding with previous owners (doublespending)

Arks: same but party can't censor, and can't rugpull deep in the past even colluding with previous owners (unless they also get 51% hashrate)

Poon-Drija channel: same but party can't rugpull/censor every few weeks if you don't re-roll-in (unless 51%)

Simple onchain L1: same but party can't rugpull using old states if you don't punish (unless 51%) At every step, security increases. But so do onboarding costs, liquidity challenges and waiting times. A smart Bitcoin wallet will automatically manage this spectrum from small pockey silver and fast transactions to big savings amounts and long-term storage. The Lightning Network will be the commont language to swap across all these security models. Another entirely orthogonal spectrum is about WHO the trusted third party is.

Could be a single guy like callebtc (practical&simple but weak security).

Or a single, reputable global federation like Liquid (easy to censor but unlikely to rugpull you).

Or a patchwork of many, small local federations like Fedis (harder to censor globally but more likely to rugpull locally).

Or the many semi-trusted peers each of us already have commercial relationships with.