When it comes to injection sites I believe almost all of the studies are being done by people who wish to see the government spend money on harm reduction research which means they set up their studies in a way to make injections sites look as good as possible by emphasizing the positive while minimizing the negative. So it comes as no surprise that the "evidence" says they work.
Interesting and much discussed theory, but do you have any evidence to support it?
Obviously confirmation bias exists, but to what extent?