Did you read about the recent sudden sharp drop in BU nodes? If not, do that. It is all you really need to know to be sure you don't want to hand responsibility over a $20 Billion network to reckless amateurs that don't review the code they crappy code they write. We can just be glad that this bug was exposed before a hard fork, not after.
BU nodes go offline. but the diversity of other nodes kept going..
however if everything was just core.. expect another 2013 event.
diversity and more then one codebase is GOOD for the network
relying on one team that has one king is no better than fiat. and that is bad.
BU, xt, classic, and many others (including some core tweaked version, will all run happily together if a dynamic event occured.
but blockstream(core) only want blockstream approved code running which if one goes down. they all go down.
oh and if there was a bug in dynamics requiring a downgrade. blocks just have to drop to below 1mb again and everything is back to as it is now.
if there was a bug in sgwit requiring a downgrade. anyone with funds on segwit keys cant spend them.. locked