I find it very disappointing that people resort to anecdote when data is available. Granted, it's a new area and there hasn't been enough time for proper longitudinal studies, but what data we do have generally supports EC's for smoking cessation, and also functioning as a replacement for young people that would likely otherwise start smoking cigarettes.
https://www.ncbi.nlm.nih.gov/pubmed/25301815
https://www.ncbi.nlm.nih.gov/pubmed/25888354
https://www.ncbi.nlm.nih.gov/pubmed/26597453
As for the safety of the vapor itself, it's a little tricky. Some studies faced a lot of criticism because the 'vaping' method did not reflect actual usage. In one protocol (unfortunately I can't recall the article) it was found they basically burned up the coil and reported combustion products as the vapor profile.
Similarly, this study appears to be reasonably executed, at least at first: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4961046/
A few takeaways from the article: no long-lived free radicals present, and short-lived radicals in the range of 100 to 1000 fold less than cigarettes. Their estimates for a typical 'vaper' would be about a 10-fold higher dose than typical air pollution exposure for a 24 hour period (200 puffs/day). This certainly represents a real potential hazard, but is also a clear improvement over cigarettes (especially because this isn't the major hazard in cigarette smoke).
However, the protocol uses 5 second 'puffs' on a 1.5 ohm coil with an absolutely pathetic 500ml/min flow rate. For reference, 500ml is the typical breath volume for a human at rest, with >10,000ml/min being typical breathing for a person at rest (https://ehjournal.biomedcentral.com/articles/10.1186/1476-06...). In actual use, a puff is usually larger than the typical breath and taken in a span of 2-4 seconds. A flow rate of 20,000-40,000 ml/min is much more representative. It's no surprise that a coil with such a low flow rate would get far hotter than what is seen in actual use. Worth noting that high flow rates make for more comfortable and pleasant vaping, so are common in actual use.
Some good work has looked at variation in vaping protocol https://www.ncbi.nlm.nih.gov/pubmed/24832759 These findings show just how drastic the differences are when you start to use higher power levels in ECs.
The data clearly show that high-power vaping likely presents a significant risk, but to date I've yet to see data that is convincingly representative of anything like normal use.