Monero promises high resistance against blockchain analysing.
Have you ever met with a tool that could break this resistance?
It does not have to have 100% success rate, I just would like to know how far the attempts could get.
You can more or less work this out for yourself. Assuming no vulnerabilities in the implementation, and assuming no such things like compromised machines, or compromised internet connection (which have nothing to do with the crypto currency protocol itself), here's the gist:
You have to assume fractional "real world" knowledge about certain transactions. That is, you can know some stuff, like Joe went to Mary's coffee shop and paid his coffee with transaction X. Joe withdrew coins from an exchange with transaction Y. And so on.
The aim of chain analysis is to use this partial information of known transaction/person links, and to complete it using the block chain itself. Here we assume that we DON'T use other information, like IP addresses, compromised computers and so on.
The problem with transparant chains like bitcoin is that this partial information is propagated, because we can follow transaction to transaction ; as such, we can learn that Joe was also involved in several other transactions, by looking at the combination of his coins and his change addresses.
A solution to this is to use mixers. In bitcoin, you have to take the initiative with others, in DASH, you have to take the initiative but it is then done by master nodes, and in monero, it is done automatically by selecting random addresses on the chain.
Mixing makes the propagation of the partial information ambiguous, but not totally void. At each mixing stage, you "dilute" your partial knowledge some more.
Now, the big difference between bitcoin and DASH on one hand, and monero on the other, is that mixing is "far and in between" with the first two (because requiring a voluntary act, and a limited set of participants. While monero does a kind of "mixing" at EVERY transaction, and selects the participants *randomly*. (strictly speaking, it isn't mixing ; but concerning the propagation of partial knowledge, the ring signature in monero and a mixer, behaves somewhat similar).
Monero has one further advantage: the ambiguity of which transaction actually happened. In a mixer, all incoming coins are really spend, and appear in the outgoing channels. With monero, the incoming coins are not necessarily spend. They can appear several times, and you don't know WHICH "mixing transaction" actually propagated them. It looks like multiple spending of the same coin in monero, but only one is "real", though you don't know which one.
All this means that with monero, you need much more "partial knowledge" in order to "complete the trace" than with bitcoin or DASH. But of course, from a certain amount of partial knowledge, chain analysis becomes moot: if you know already ALMOST everything, then the little it can buy you is not so important. Chain analysis is interesting when only a little bit of knowledge can learn you a lot.
If I have to know 90% of Joe's transactions before the monero chain tells me the 10% that remains, that's much less interesting than when with 10% of Joe's transactions and the bitcoin block chain, I can find out the remaining 90%.