Sensitivity must be applied to respect the integrity of the actors involved. What is and is not revealed must be a matter of aligning the findings with the research purposes. Any auto-ethnographic narrative must ultimately be an approximation to the truth, since it is presented through a very personal lens. Regarding 'what to tell and what to hide', research validity will increase where transparency is enhanced. There will be issues to mask for a range of reasons, including ethical reasons and the protection of the actors involved. Acknowledging that issues exist in a sensitive way can provide the reader with an assurance that these difficult issues have been considered rather than concealed.
Dear Ph. D., very good question. I have not seen research on this subject, in my opinion the fundamental problem is to do research without affecting the interests of other people and society. The principle that should govern relations with other individuals is the following: It is not about subordinating your interests to those of society, it is about inserting yourself into society without affecting others. So you should continue this investigation. Someone said that respecting the rights of others is to consider the moral principles that govern social development.
David Calvey in "Covert Research" (2017) writes about the ethics of covered research in general and also dedicates a chapter to autoethnography. Maybe reading it would help.
But I am afraid that there is no general answer to your question which is always very much depending on context. In my opinion, the most general answer possible is this: thoroughly think about if and how your research may harm those participating and then change the presentation in a way that it most likely does not. For instance, I write an ethnography drawing on research in the shipping industry. In the presentation of my results, I omit certain practices with which workers resisted work pressure by e.g. hiding or slowing down processes, since detailed knowledge about those practices could be used by managers to develop counter-strategies to improve the utilization of their labour power.
Juan Manuel Montero Peña, I think that social research, if it is relevant at all, necessarily affects interests of people and society, otherwise it would be a useless endeavour. In my opinion, the question must be whose interests and how it affects those.