Note that facts can widely vary, depending on the source. We have reported data, which is a normalized version of it. That makes it reliable to use in the UI and is part of the API. See #35619 as well for how a kernel_version fact was added and then consumed in #35622.

Thank you for your response. However, we would like to merge PR #35667 by the end of the next week, and I think your resolve would need someone else to do it. Is the way I approach it not suitable?

No, facts are too different. Depending on the fact source it can be reported with a very different name and it's very unpredictable. We really need to normalize it into reported data first.

Sure, thank you for your help.