That's the nature of drug testing though. By it's very definition, it can only be a reactive process. You can't ban a substance as performance enhancing till it can be proved that 1. it's being used on a big enough scale to make financial sense to investigate it, 2. it's actually performance enhancing and 3. it can be detected in a highly reliable fashion.
That's a PhD level project right there. Four years. For one substance. By the time the investigation is complete, everyone might have moved onto something else because it's more effective / cheaper / more available / easier to mask / breaks down into something legal, so banning that substance is basically pointless.
And don't even get me started on how these substances can effect each other to produce a completely different compound.