I'm confused by the diametrically opposite views.
My answer will attempt to explain why you may see "diametrically opposed" views on these types of subjects.
For a summary of evidence on ED and male fertility see answer: Can cycling adversely affect male fertility?
Quality of data
The strongest scientific inferences come from studies where you can randomly assign treatments to your experimental units (i.e. people in this case). The randomization is key as there may be all sorts factors (some we may know about, others we may not) that may impact the measured response (e.g., prevalence of erectile dysfunction). Randomization will, on average, even out these effects.
The problem with epidemiological studies - studies of health and disease in human populations - is that it is generally not possible to randomize treatments, instead researchers rely on "natural experiments" to do the treatment randomization for them. For example you cannot go out and gather a random selection of people from your population then assign them a treatment such as cycling for the next 6 months at a rate of 0, 5, 10, 20 hrs a week, so that you can measure the rates of erectile dysfunction (ED) or changes in sperm count.
Rather you must rely on people who have essentially self-selected these treatments. Are there other factors associated with whether or not a person may be chose to ride 0, 5, 10, or 20 hrs per week? You bet! Could those factors also be linked to ED or sperm count - sure. If we know exactly what factors could be influencing the response we can potentially control for these statistically. The problem is that their may also be factors we simply don't know about, termed lurking variables, and there are numerous examples where they have mislead interpretation until discovered. A simple example is that shark attacks are correlated with ice cream sales - sharks love ice cream!
Furthermore, you also rely on subject self-reporting, which can result in issues such as reporting bias. For example, ED has been a long discussed issue in cycling circles. As a researcher interested in ED and long distance extreme cycling you survey a bunch of riders after a long event. Then ask them to self diagnose these issues. The fact you (as a subject) have heard about these issues before, and now you have a researcher asking pointed questions can bias your response. That time you were thoroughly exhausted and just wanted to sleep could get categorized as impotence (because you are concerned), when in fact it might just have been exhaustion. We don't know because most of us are not trained physicians who can accurately diagnose these issues.
All of these different factors can result in spurious results (a "statistically significant" effect that is essentially a fluke) in studies, especially small studies. Since this is a "sexy" topic any study finding an effect (spurious or otherwise) will quickly make the media rounds, which will quickly make the facebook rounds and have people talking. Once a "fact" is out there it can take a long time to repeal (don't get me started on vaccination - the single study often quoted as evidence against vaccination was long shown to be a fraud study).
These factors can quickly lead to diametrically opposed views getting air time and the general public being confused. For a great overview of this see John Oliver's rant on science and science reporting.
What can you do
- View any single study with great skepticism - it could be flawed or the results could apply under a very strict set of conditions.
- Look at the effect size - some times it may be significant, but the size of the effect is small.
- Population averages do not necessarily translate to individuals (aka ecological fallacy; (e.g., the population may ride BSO with horribly uncomfortable saddles, that aren't set up correctly)
- Finally, look for "meta" studies that are over viewing the current state of knowledge looking for the weight of evidence. For example if 20 studies exist and all point to a large ED then I would be pretty damn concerned.
- The other related answer (Can cycling adversely affect male fertility?) does have one such study (Leibovitch and Mor, 2005), that suggests:
Urologists should be aware that bicycling is a potential and not an infrequent cause of a variety of urological and andrological disorders caused by overuse injuries affecting the genitourinary system.
- Unfortunately the full article is behind a pay wall so I have gotten my hands on it yet.