The bigger money is more risk-averse.
The development team structure is an area of risk.
Less confidence can be placed in an anonymous party than in identified parties. Open-source code-review, if thorough, can dispell some but not all of this uncertainty.
The relative market caps of XMR and BBR are pretty well explained by the difference in team structure.
To make this point speculative, consider that CZ could identify himself at any time, or even just disavow the Sabelnikov identity, or more broadly, participating in bytecoin.
If he did, money flow to BBR might reduce moneyflow to XMR.
The market priced this so that the XMR-invest, BBR-hedge was nicely asymmetric, but now SuperNET has pumped BBR. I consider that the hedge has paid off. Thus I tend to short the hedge when it is overbought. But when it all blows over, I expect the cheap hedge to come back, and I plan to make more use of it in my portfolio.
It was in this context that I provided these points:
What I don't get is the evidence / proof that he is / was the ringleader in the original scam. Which I don't think there is any proof - only speculation from Monero supporters?
I am definitely a monero supporter. I don't tolerate misinformation, and when I am wrong I try to make sure no one was deceived by my errors. I've never met anyone who didn't suffer from some degree of confirmation bias, however.
There is some reason to believe he is Andrey Sabelnikov. The code style is arguably similar. That may seem weak, but I know from my professional work that these things can be quite accurate. I haven't done sufficient work to measure the accuracy of this one with error bars.
In general, he demonstrated a much higher level of mastery of the codebase much faster than anyone else. There's no evidence that this results from the sort of 30-point IQ differential that would be required if he had no prior exposure. In fact, there are both population statistics and anecdotal evidences to refute that hypothesis. The obvious remaining adequate explanation is extensive prior exposure.
Few are really conclusively convinced by these arguments.
Do we know the code actually makes transactions anonymous? Has there been enough review to determine that?
Yes. The probabilities are well-defined, and computable in closed form.
EDIT: This applies to the math, not to the implementation. Substantial review has been conducted, but much more is required.