I think the answer depends on (the perception of) the clientele of the forum.
I use clientele as an abbreviation for class of people to be served. Many of the clientele are professional researchers (or soon to be professional) who have learned enough of the craft of research that most would not think of asking any question about the process itself on this forum. They see MathOverflow as a research tool for their specific focussed question. (They may see MathOverflow in other ways too, but usually research is so core to their existence that the word research itself seems redundant, or like excess baggage; they see no need for the tag.)
Many of the clientele are people who want to move toward research in mathematics. They may not be professional mathematicians, but they have or see the need for mathematics at more depth than appears at an everyday level. Some of these clientele are researchers in a different profession, and may not be familiar with the characteristics involved in conducting and presenting research in mathematics. They may have questions about how to do this. I can think of no other forum (not even Academia.stackexchange) equipped to handle such questions. For example, what is a good way to set up an Atlas of relational structures for use by mathematicians in other fields, especially if (for some unknown reason) one restricts relations to be ternary or of higher arity? Such a query on Academia would quickly be migrated here, or at least the migration would be attempted.
There are (I believe) clientele of this forum outside of those two categories. (I know at least one such member.) They may not be adept at research, or at their field of study, or even at posing well crafted questions for MathOverflow. They still can contribute something, and they may have enlightening questions having to do with the process of research. I think an appropriate question for this forum would be how to determine what work is publishable where. In particular, if you have a result and you are doing a literature search, how do you announce it and get help on the search to determine how novel the result is? Let me ask this general question on MathOverflow. Gee, what tag should I use?
Finally, there is the non posting clientele. They may not ask any questions or make posts of any kind. However, they use this body of knowledge regularly to answer questions. If some of the questions pertain to describing the process, they may deserve a research tag.
If you want to fragment the tag into specific pieces of the process, that might be a good idea. (It might be a bad idea.). If you want to remake the tag to emphasize that this forum is about mathematics, go ahead and call it mathematical-research. I think removing the tag will do a disservice to all but the first set of clientele described above.
Gerhard "In A Class By Himself?" Paseman, 2020.06.13.