Yes I can. It is a fundamental taxonomy of possible workarounds to the FLP impossibility result. Read my whitepaper when it is released (although I don't think you will understand it but it is written to be comprehensible for those with reasonable analytical skills).

What @jaekwon (Cosmos/Tendermint) and others may not realize is that the fundamental limitations of Byzantine agreement can't be structured around with protocol. It is a fundamental limitation (due to physics of asynchrony). So they can obfuscate as much as they want with heapings of protocol complexity, so as to hide the fundamentals from themselves. But I understand the essence of genius.

So Casper is hiding the bonding collateral in the notion of betting from collateral, but they won't escape from the fundamental limitation which is that blame can't be objective (it is always ambiguous) in Byzantine agreement. The only alternative to Byzantine agreement is probabilitistic, asymptotic (e.g. PoW and my design which is not PoW but something new). But probabilitistic, asymptotic can't assign blame to malfeasance either (for example you can't prove that Bitcoin miners are censoring transactions from blockchain data objectively, you need external social information which is not objective).

I will double-check my logic on this again when I do the comprehensive re-read of my paper. Will report back here, if I find any flaw in my analysis.