I've been out of the industry for many (many) years, but here's a bit on how problems can creep in with well fracing. Background: I was in well testing, and have been to many a fracked well during/before/after the frac was performed.
We'll start with a well that is drilled, cased (casing is a solid-steel pipe all or some of the depth of the well, used to keep "stuff from coming in, or stuff from coming out"), and perforated (holes shot through the casing with explosives). Typically all of this work is done by contractors. The oil company leases the ground being drilled on. Everything else... the oil rig, the drill pipe, the workers... all of it belongs to other subcontractors. One "company man" from the actual oil company sits in a trailer on site to monitor the work. This involves a lot of waiting for someone to ask a question, and playing solitaire.
The oil company now has an outside contractor come in to frac the well. Literally, this is the next morning after the perf job if possible, because a rig costs tens of thousands of dollars a day to sit there and wait. Over a dozen big trucks come in at the crack of dawn, and link up so that over over the day, viscous, proprietary-formula fluids can be pumpted into the well to induce cracks in the formation from the overpressure of the pumping. Then a proprietary "breaker" fluid is injected to make the original goop less viscous, and to make it drop the sand embedded in it to hold these cracks open. The former goop, now runny (fingers crossed), will flow out as the well produces. The trucks are out of there the second they're done; *they* cost money sitting around too, and they're probably off to another job the next day.
-inc soapbox
My personal, biased opinion of the disconnect with fracking, the industry, and its effects, is that there is a science problem, and an accountability problem.
Scientifically, there are a number of wonderful calculations that tell us how we're effecting events inside the well. These models tend to assume an understanding of the various strata and depositions drilled through, and can easily confuse the ability to make a model match an event, with the ability to understand the mechanics of an event. This leads to an environment where current perceptions of the industry and the confidence/ego of the simulation's creator are the deciding factors. Since much of this science has migrated out of the oil companies and into the contractors over the years (or to contractor-supported academics), the operators now base their knowledge on what the contractors say is correct (this is an oversimplification, but overall I feel it is correct).
On accountability: trade secret formulas mean we have no idea what is pumped in the well. The "in and out" nature of the fracking process means that crews who perform the work have little exposure to the site, and no connection to followup on the effects of their work. Oil companies serve as the face of the project to the land owner, but have outsourced all the science to the contractors, and are defending work they understand based on the explanation of a salesperson to a client.
-rem soapbox
The above problems do not at all prove that fracking is bad, or good. They do, however, create a disconnect, making it hard to develop a cohesive picture of what is going on, good or bad.
To put it into (hopefully?) a useful tech metaphor, the contractors make the computers, and the oil company sells them to people. People complain to the oil companies that some of these computers are terrible. The oil company naturally says "oh no, we only sell good computers," and runs to the contractor. The contractor tells the oil company, "No, they're great, look at these schematics. Those people are outliers due to blah blah blah." So the oil company gives those people their money back, and makes them promise not to bad mouth the computers they're selling. Repeat as needed, until the evidence of problems with the computers is so great that the oil company cannot ignore the truth any longer, and starts selling someone else's computers at great cost and effort. Because... those computers will have *great* schematics.
