The article isn't science ... the first comment courtesy of dansus that follows the article is science :
"Propylene Oxide
They did experiments with pure PG and pure glycerin (VG) and found glycidol in the PG but not in the VG. The concentrations found seem likely to be insignificant though. NIOSH set a safe workplace exposure for PO at 100ppm for an 8 hour day, which translates to a maximum of 998 mg inhaled in a day. The worst e-liquid offender in this study contained 6.7 mg/mL, so assuming every single mL of e-liquid was inhaled, you'd still be in the NIOSH safe zone if you vaped about 745 mL of the worst e-liquid offender from this study per week.
PO is in the same carcinogen class as aloe vera, pickled vegetables, or working as a carpenter or a dry cleaner. It's also highly flammable, so if you notice that your e-liquid will easily ignite it's probably best to switch brands.
Glycidol
Glycidol was found in the vapor produced from both e-cigarette tanks, but at levels about 8 times higher in the dry-puffing CE4 than in the Aerotank, suggesting this is not something just in the e-liquid itself, but something being generated by heat.
Glycidol is in the same carcinogen class as drinking hot beverages, eating red meat, or being a barber or hairdresser. Between OSHA and NIOSH, the maximum allowable workplace exposure to glycidol is 264 to 528 milligrams per day. To reach the lower 264 milligrams of glycidol exposure using the Aerotank, you'd have to vape about 1.2 lbs of e-liquid per day. In the dry-burning CE4, 264 milligrams of glycidol would take about 0.15 lbs, which is about 68 or so milliliters of e-liquid per day, and then you're still only right at the safe level of the more cautious organization (these calculations are all rough and assume 50/50 VG/PG e-liquid is around the same density as water). So I think we're pretty fine on this unless we're vaping a pound of e-liquid per day or vaping nothing but dry hits.
Formaldehyde
25 watts in a 3rd gen Kanger Subtank produces 0.34 µg of formaldehyde per puff, compared to the 45-55 µg produced by the CE4 in this study, which is more than 100 times higher. When researching e-cigarettes, the production of high levels of formaldehyde is certainly evidence that there is something wrong with your methodology.