This is a follow-up to my question on the mymonero backend here
Just had a quick look at the openmonero implementation of the mymonero backend (pretty well readable btw given that I'm not familiar with the monero code base). Looks to me like the main point the mymonero backend adds is that the blockchain scanning, which in monero-wallet-cli is part of the wallet is put into the mymonero backend, which makes it kind of an intermediate layer between the regular full node in the background and the mymonero app. Is that to make the mymonero app even less computationally intensive than monero-wallet-cli and monero-wallet-gui? Or is it more in order to keep the bandwidth usage of the mymonero app low, because only the transactions for my address are sent from the mymonero backend to the app, instead of the whole blockchain?