I understand and I do agree with all of your examples.
Where I don't agree is that my primary source is not based on consensus from beginners, that's all. The above example on the survey on betta vs guppy is based on their actual user experiences rather than what they heard. If one species break the other one apart, for them, that combo isn't working. I don't think that example is quite the same as the ones you mentioned here...
But actual user experiences can be fatally flawed as well. Again, look at my above example about pH crashes. In many people's "actual user experiences" their fish died because of a change in pH. Even many LFS shop employees would tell their customers that. However, once I read some actual scientific research on the subject, I started to question it, and am now convinced that pH itself is a very minor disturbance for most fish. I can cite the research papers on which I have based this opinion. But, if you just continued to go by user experience, beginner or experienced, it was very unlikely that the knowledge would have just come out.
Betta versus guppy is the same way. What are the actual numbers? Do you have 5 anecdotes or 500? And the word anecdote is used very conscientiously, because unless there is some sort of actual study done with controls and the like, all you have are anecdote. And anecdote is only very rarely better than having no information at all. That's where having some scientific method becomes so very important. With anecdote, all you have is 1 situation that occurred one time without any controls in place. With a study, you can figure out if what happened has statistical significance or not. That is, whether it is an actual trend or if it was just random luck what happened.
Lastly, and this is something that I should have emphasized more above, but science and facts are not a democracy. Just because a lot of people hold one point of view, does not mean that that reflects on reality. The facts are not determined by popular support, or by consensus, or any other word for it. That's the true power behind science. When it is done right, in an objective, unbiased manner, science gives answers completely independent of any preconceived notions at all. Now, that idealism is rarely realized -- the real word is rarely perfect enough to allow us to study exactly what we want to study perfectly cleanly -- but the main point is again that it is not how popular or how often a point of view is expressed that determines its rightness. It is how well supported it is by the evidence.
So, again, if you support the advice with good evidence, then there isn't an issue. If it is just the most popular or prevalent point of view, then there is a significant risk that all you have is the popular point of view, not the actual facts.