OK, people have been saying you'd need to control the entire network with blinding in place, or at least an unrealistic portion.
what weighting do you think is realistic to perform extrapolation and gain sufficient control?
Unfortunately it's hard to quantify without a mathematical expression of the model, especially without a great deal of familiarity with the inner-workings of the system at present. It also stands to reason that the level of analysis required to properly document and model the system would require a very motivated (money, inclination, whatever) cryptographer, which I am not (as in not really a cryptographer, nor particularly inclined:) )
OK, well we are frequently hearing 'no GUI wallet for XMR is no problem because we have mymonero' yet this is centralised and subject to attack. That's all really, and if this is a temporary issue then fair enough.
Oh I totally get it now! Sure, as a stop-gap solution it presents a risk to a motivated attacker. To be honest, and this is not to be overly pragmatic, we haven't even implemented the changes recommended in
MRL-0004 yet, so Monero isn't yet at a level where I would feel comfortable recommending it to someone who transactional privacy is a life-or-death scenario. Once we've implemented the MRL4 changes, and have our hybrid i2p/ip layer running, that will decrease the remaining risk areas (as we see them) to "negligible" levels. Although security and privacy is an ongoing process, so we're never going to stop researching and theorising attacks against Monero, and doing what is necessary to protect our users.
At some stage, when I get the time for it, I'm going to build a very aggressive and malicious Monero testnet node (ala
Chaos Monkey), and permanently run an army of them on testnet, to force us to build more anti-fragility into Monero.