Maybe I can tolerate it now but I’m just so scared to try and cause a pain flare.
I would feel comfortable recommending a trial of a small glass of green tea, one tea bag, brewed in 250mL/8 oz water (that was literally just at a boil), for as close to 3 minutes as possible (watch the clock). This should produce a drink with around 20mg caffeine, and you can drink half of it and watch for any symptoms, which should be mild if any.
This study gives the rather unhelpful information that 4.5mg/kg caffeine is not good. For a 60kg person that is 270mg caffeine. Uh, pretty much everyone would be busting for a piss after that much...
This slightly more helpful study does seem to show a dose-response relationship. They found drinking up to 1 cola-type beverage a day has seemingly no effect (odds ratio for lower urinary symptoms of 1.01, meaning they were 1.01 times as likely to experience some negative effect as a control without caffeine), more than 1 had a small effect (OR 1.12), up to 1 diet cola (which have more caffeine in general) had a very small effect (OR 1.07) and more than 1 a small effect as well (OR 1.18). They didn't seperate tea by black/green, but both up to 1 cup and more than 1 cup had similar moderate effects (OR 1.31 and 1.34). Coffee was the worst, 1 to 2 cups had OR 1.41, more than 2 cups, OR 2.09.
Curiously they found citrus juice had a protective effect, 0.4-1

z serving a day made symptoms 2/3 as likely (OR 0.67) and more than 1 serving made them 1/2 as likely (OR 0.50).
So, if you do caffeine, keep doses as small and infrequent as you can, drink some orange juice alongside it.
Sekio why do you not think caffeine is used to bind the fake blues? And you think legit binding agents are more likely?
It's possible, I've just not personally seen it. It would make sense if the pills are completely volatile and leave little or no residue though, as the vast majority of pill fillers will burn, char, or just melt instead of evaporate/sublime, but only caffeine is readily available, will "smoke" cleanly without residue, and is also water soluble if heated. It is a traditional cut for heroin for this reason.
I can't actually say what binding agents some clandestine manufacturer would use. It would depend on desired qualities, availability, the local customs, the "recipe" the "manufacturer" uses, etc.
For instance there are 8mg hydromorphone hydrochloride tablets that are made with nothing but lactose as filler, which leads them to being totally soluble with heat, leading to safer IV and nasal use.
I figure it should be pretty easy to test these for caffeine with a reagent test….or any analytical labs you know of to check databases where these pills have been analyzed?
I believe it turns brown with Mandelin reagent, and can be separated using TLC and stained with it to verify, if you are willing to go that far.
The problem with analyzing fake pills is, how can you tell which fakes they are exactly, if there are multiple types of fakes with the same imprint and different "recipes" you'd be left to guess which was yours...
Or, if the pills are mostly caffeine as filler, just measure the melting point, perhaps.
Caffeine is more expensive than mentioned binders and at production levels such as with fake blues it can be significant sum of money.
Millipore Sigma has 25kg for C$65 a kilo, another web supplier has 50kg for $32/kg.
And in larger amounts from more specific suppliers it is even cheaper.
This is not an insurmountable expense for a product that is sold, at street level, anywhere from $1 to $20 for a maybe 300mg pill. (At $32/kg, the cost works out to a whole 1 cent per pill.)