It is just a definition
If the definition is useful really depends on what you want to model.
The author of the SEP article does not know what you want to do, so he includes a reasonable amount of hopefully useful definitions.
Usually, it doesn't make sense to be reflexive, the article itself says so.
There are some basic properties which can naturally be imposed on instant-based flows of time. It is usually required to be a strict partial ordering, that is, an irreflexive and transitive relation
From SEP, my emphasis.
But as an example, if you want to describe a repeating process, circular time is a natural way to do so. Since transitivitiy still holds, every point precedes itself. Or think of a process that might branch, and possibly has loops. Then points that "precede themselve" are exactly those which sit in a loop.