Couldn't agree more. As I read the title I couldn't help but think: "The OP just opened Pandora's box"This debate will rage on for years to come.
Look, at the end of the day, you are dealing with sound pressure levels. Your eardrum doesn't care whether the source is a set of speakers, environmental sound, hearing aids, or headphones, etc. The SPL at your eardrum is all that matters. In other words, 60 dBA at the eardrum via headphones is no different to 60 dBA from any other sound source like conversational sound, for example. The sound wave hitting your ear would be scientifically identical and would carry the same amount of energy (ignoring infra and ultrasonic frequencies, of course).
There have been many studies conducted where special microphones were used to measure sound pressure levels at the eardrum. I'm not aware of any study that has suggested that equivalent low-level waveforms from headphones are different or inherently dangerous to human hearing. On a logical level, this makes sense as we are just following the laws of physics. If headphones are dangerous then so are hearing aids and white noise generators, that is unless they warp the space-time continuum somehow
The real risk with headphones is ear fatigue from overuse and a phenomenon whereby people can't help but slowly increase the volume over time without realising it. As the brain normalises the incoming sound it's really easy to get into dangerous territory unknowingly, and background noise can play a big part in this process. This is why both passive and active noise cancelling headphones were invented.
There is nothing out there that can validate the idea that headphones are inherently dangerous at safe volumes. @Bartoli is correct when he said it's flawed to use data exclusively from a site full of tinnitus sufferers. There is no control for this which makes any conclusion meaningless and anyone who attempts to interpret such data either way is likely only confirming their own bias. This is why the scientific method is so rigorous, it has to remove all potential for human bias which is rife in all experimentation.
If one feels more at ease by never using headphones again then I'd say that's a good course of action, but for someone to claim outright that headphones are inherently dangerous at safe levels (and time exposures) is a bit misguided, I feel. There is no evidence to support such an idea.
Statistical analysis is a minefield. There is no way that one could conclude that headphone usage was the cause of a worsening. What else was that person doing? If they all ate cheese as well, then maybe it was that? Did they all wear shoes that day? How do we know the shoes didn't cause it? Did they take antibiotics? Etc, etc.
This leads me to something I've already mentioned on here but I'll say it again. I was at a lecture once where the topic of statistical data and it's interpretation/analysis was spoken about at length. I also have a book about this which is a good read (forgot the title at the moment). There are so many facets to this that I couldn't possibly fit them all into a paragraph, but I think it's something worth knowing. One of the main points that sticks out, which he was very clear about, is when we automatically attribute a cause to a certain effect, and believe it to be true. Even when the evidence overwhelmingly suggests it. He simplified this with a demonstration that involved people who had experienced a yellowing of their teeth, that he labelled (A), and people with lung cancer that he labelled (C). He stated that the data would suggest - with a very high probability - that anyone with yellow teeth (A) would have a significantly higher chance of obtaining lung cancer than those with whiter teeth. This is where data can be deceiving and why its analysis is so difficult, because different lenses can be applied and truth can still be found. At this point he introduced smoking and labelled it (B), and only then does the scenario begin to make more sense. The yellowing of the teeth (A) becomes almost irrelevant once you add in smoking (B). Because it's then that you can see that smoking (B) causes yellow teeth (A), but it can also lead to lung cancer (C). The link between (A) and (C) is no longer important, but is still true. Smoking (B) is the real cause, but if (B) is not known or understood then you wouldn't have the information required to understand what is really happening.
This is why human biases are problematic.
I kind of knew what I was getting myself into when replying, but I still think correct information should be provided so the OP can make an informed decision. So I said what the heck and dove in. Even without a black belt in tinnitus.