Improve node lifecycle handling

[jvsena42]

Problem:

the node once broken (trying to stop but never succeeding) won't fix itself even if I wipe app and create new wallet

had to close the notification / un+re-install app to get it to work

not sure if it had to do with the backend issues, I'll let you know
also there's a very crazy issue which will get fixed when Ben ports the iOS implementation to Android:
after RBF, app is always loading and never completing/working good; because it keeps having issue as it can't find some TXs via Esplora queries in AddressChecker
and one thing that would be nice:
if node is broken / can't start; it doesn't auto-heal when the backend(electrum) is back online

Suggestion

Something like this might improve QoL:
when not started properly, whatever it does based on waiting for stuff; hard reset and retry to start from scratch in Lifecycle.Event.ON_START 
by hard reset I mean kill/ destruct via Rust, without waiting in a coroutine's fashion; or cancelAndJoin if coroutines are preferred
or, maybe we can optimise stop to not wait so long, as it does now at times;
I guess this is the real issue
the fact that usually stop() is not optimal; it waits while it shouldn't