Things I do or would rather not draw attention to.
I'm an entomologist who has worked with various agricultural pests at a university. I'm mainly an ecologist trained in integrated pest management, so I've dealt with toxicology (mainly effects of insecticides on pests and beneficial insects), biological control, population dynamics, etc. I normally frequent agricultural and entomological related topics because of that. As a scientist, I'm especially interested in how scientific findings reach and are interpreted by the general public. With that, my main interest on Wikipedia is seeing that scientific articles match up with the published literature and are appropriately summarized for general readers.
With COI in mind, I don't edit on specific topics that I research directly and follow the general guidelines for academics. That means that while I may edit the page of a pest I've worked on, I won't be editing content pertaining to my research topics specifically (or citing my own studies for that instance). All my work is pretty niche though, so I functionally won't have a COI in any topic on Wikipedia (especially while not citing myself) unless someone decides a to write a BLP page about me (which I hope not). It seems to come up often in articles I edit, but I don't receive funding from private companies (i.e., pesticide companies), nor do I have any personal or financial connection to pesticides or companies marketing them in general as a university researcher.
I also follow guidelines for concise and accurate writing when editing. If it seems like I have removed content, it is very likely the exact same content was summarized earlier in the article and was redundant, or a couple deleted sentences could easily be summarized into one. This is to improve the general readability of articles that sometimes get too verbose for general reading. My approach to sourcing very closely follows WP:MEDRS, which does a good job explaining scientific literature in general. Feel free to leave a message on my talk page if you have any question about edits or articles I'm working on.
I take a lot of focus on sourcing, especially in scientific topics. In nearly all cases, we want to avoid using primary sources per various policies and guidelines. In scientific literature, primary research articles do undergo a round a peer-review by one to three scientists in the field. However, as taught in most introductory science sources, the purpose of publication is not to accept an accepted publication as truth. Instead, the purpose of publication is to say it passed the first round of inspection and is sound enough to be vetted by a larger scientific audience. At this point, such sources are not particularly useful for Wikipedia because their intent is to be read by other scientists in the field who are also assessing the methodology of the study, its experimental design, statistical tests used, and if the conclusions actually are supported by the data.
We cannot engage in original research to assess all that as Wikipedia editors, even as an expert editor, so we instead rely on secondary sources (i.e. review articles) that cite the primary study to vet it for us per WP:WEIGHT. For that reason, including content about the findings of a primary study by only citing that study is not going to be appropriate for a Wikipedia article. In Wikipedia, some scientist editors may want to cite primary studies like they do in their own research publications, while other editors may not be familiar with the framework of scientific publication and want to include primary studies claims at face value. That can be a source of contention and confusion, hence my specifically mentioning why this is problematic above.
With an understanding of how scientific publication is intended to work, there are rare instances where a primary source can be appropriate. In a non-controversial topic where review articles or other secondary sources are not published frequently, the introduction section of a primary source can serve as a mini-literature review of other sources cited. These introductions tend to be limited in scope to the specific experiment and likely won't be comprehensive, so they should not be heavily relied upon except for very basic facts. This can include basic biology such as the life cycle of an organism, habitat, etc. Such sources should be avoided in controversial topics though, as particular viewpoints can be cherry-picked from primary research articles that manage to get published but are not cited or taken seriously by the larger scientific community. However, the results and conclusions of the paper are original interpretations of data, and would require a secondary source since we as editors cannot evaluate the validity of the claims made.