There are still a few journals that do not accept submission of manuscripts that have been published on a preprint server. Besides, is there any other cons of publishing preprints?
You should avoid telling the other researchers about the details of anyone of your papers until it has been published and seeing your name by yourself.
Why you let others know about your insights and methodologies before publication? You may say that I am somewhat old-fashioned, but I have a different perspective for uploading any preprint anywhere before it has been published by your name. So, my advice is to not put your research anywhere until it is published.
Your manuscript may be copied and then published by others before you can do that. This stealing of your paper might be happening. You must wait until the paper is accepted and then published by that journal. Then, upload that research item on any platform you wish.
A journal may have automated plagiarism software to check the paper before admitting it to the reviewing process. There are chances that your paper can get a rejection at any point. Thus, to avoid this problem. Publish the preprint after you got "Your paper is ACCEPTED".
Academic publishing remains a competitive process. If someone else has recently published a paper very similar to mine, mine is less likely to be accepted.
So although I may share my topic, I prefer to keep my methodology, findings, and discussion private, until published in a peer-reviewed journal.
Papers in certain fields such as mathematics are so specialized that they can take up to 4 years to get published. Having them on a preprint server means that you can immediately lay claim to the fame of the method, if new, and be cited already. There’s a saying in English that “great minds think alike.” The point is, holding your paper back in the hope of perfecting it for the “perfect journal“ or time may actually cost you because, rest assured, one mind among the thousands out there may be working on the same exact problem. The history of science abounds with such examples. Then, of course, having them on a preprint server means that your work receives critical reviews before actual publication. The fears of being plagiarized aren’t really founded because once your work is preprinted, it is dated and digitally identified with a DOI. So, proving precedence is not the problem. Technically, you can use that as proof against plagiarism. How easy is it to fight the plagiarizer in practice? that’s the question!