Proof? To me it seems like the changes in
https://github.com/stellar/stellard/commit/067d7158720331937fc782cbb230e8d422cd7341 (especially "Consider there is consensus when we detect that we've fallen behind") which are exclusive to Stellar are potentially to blame, not their validation network topology (validators that were operated by one entitiy - SDF - started to disagree with each other, even though they likely had identical UNLs).
I also was saddened to see that no one noticed the dissonance in the 'temporary' solution of converting to a centralized model: If a system can be converted by some loss correcting central bank into a centralized system ... can we really say it was ever decentralized in the first place?
If that "bank" is system relevant, maybe...? Consider that one day a large majority of Bitcoin exchanges and service providers (BitPay, Coinbase...) decide to only accept Bitcoins sent on their centrally mined block chain. Either they all remove themselves immediately from the Bitcoin community or the Bitcoin community has to move over to their centralized solution. While Bitcoin's community consists of lots of people with lots of different ideas, that's something that is not very likely to "fly"... Stellar on the other hand consists mainly of fake accounts that were used to grab facebook giveaways. Also not a lot of people seem competent enough to run their own nodes (it is not incentivized after all, compared to Bitcoin where at least miners need to run full nodes). In practice, the SDF likely is the only major player (main hosted wallet that contacts SDF hosted nodes by default) in that ecosystem, so it is not that hard to (re)take control over the whole network.