The outcome depends a *lot* on just what the fab labs are capable of--if all they can do is plastic, they'll make a difference. If they can do electronics as well, it's a bigger difference. If they can do complex organics (food? drugs? viruses?) it's huge. And if they can do living things, life will get genuinely weird.
Another question: will fab labbed items tend to be cheaper than manufactured items, or mostly just more convenient/ more customizable?
In any case, I'd expect energy, infrastructure, and personal services (have you read Heinlein's "We Also Walk Dogs"?--it's about a company that offers all sorts of legal services, a sort of super-concierge) to all survive fab labs very nicely. It's not obvious which would be best. I'd expect real estate prices to rise if people are spending less of their money on stuff until and unless tech reaches the point where there's no huge advantage to living in or near a high density area.
Here's a non-standard suspect in re medical waste: From _The Lost Art of Healing_ by Bernard Lown, there's the idea that hasty examinations are hideously wasteful. The first complaint a patient mentions is frequently not the real problem, but if the examination is too brief, the patient not only fails to get treated for the actual ailment, but may suffer side effects while being treated for the wrong thing. And the patient or the government or the insurance company or somebody pays for it all, too.