Maybe because all this time nobody made an analysis of the Bitcoin transaction graph which prompted an investigation into ways to determine common ownership of addresses 
. So you can thank Adi and Dorit for that.
More like: the fact that their analysis didn't turn that up further confirms how shallow and ineffective the research was. If only there were real research happening it might have been discovered more quickly! (Of course, many other people did do things with the transaction graph without noticing this too see for example blockchain.info's taint tracking service. So they were hardly alone in not noticing it). So you can thank Adi and Dorit for that.