Have someone test it, it is not even remotely plausible to buy your own equipment, nor run a GC-MS, they don't just come with instruction manuals.
They are typically built to order and take months to receive and shipping a several hundred pound machine isn't cheap. Plus all of the standards you have to have to calibrate them, as well as the gas they operate on isn't very cheap, plus the chemicals it takes to extract the compound into an analytically viable sample. All of this also requires space and permits/certifications to set up, operate, and store. Takes about a month to get all the software working with the standards dialed in. Plus the cost of the machines is ~250k for a new one.
Labs, regardless of what it is they are testing are audited, by both internal and external entities, for safety, procedure, and data integrity. The cardinal sin in a lab is to submit knowingly false data and it can get a lab shut down extremely quickly. Plus the fact that most lab workers sign a confidence statement that basically makes any bad data the fault of the analyst-i.e. customers who have been cheated don't sue the company, they sue you and you end up in prison, and when you get out of prison plan on a new career, bad data will get you blackballed from a very large majority of any available laboratory jobs.
In short I am sure that the labs in operation submit quality data for their samples and are a much cheaper solution, I am sure they are accurate in their analysis. If you have questions about their accuracy most of them have certifications that they have to make available to the customer and are most likely visible on their website.
http://fullspectrumlabs.com/validation/ is the method validation for Full Spectrum Labs, basically a written and approved method for the determination of the cannabinoid chemicals found in marijuana as well as the numbers to back it up, just grabbed that at a few second perusal of their site, I am sure their accreditation is there somewhere as well, by the way looking through their method validation these guys are on top of things. As to accuracy of the GC-MS, to borrow from Samuel L Jackson in Jackie Brown a bit-"When you absolutely have to analyze every fuckin chemical in the room...accept no substitute." These machines measure in ppt or parts per trillion, they are extremely sensitive and extremely accurate. At my lab you aren't even allowed to wear your labcoat from outside the GC lab into the GC lab because you can contaminate blanks just by being in the room with chemicals they analyze for on your coat across the room. Samples for GC analysis are not allowed to leave the GC lab due to the ability to pick up contaminants just from being around them. Plus the machines are calibrated whenever your quality control standards fail and the samples are rerun until you get passing results.
Sorry for the wall of text, just something I have a bit of experience with.