the i partition is part of the pseudo disklabel that openbsd generates for
disks, so that there's always a disklabel even if it is in memory (or "in-core",
as the manual states).
other systems don't care about our disklabel as it's simply an additional layer of partitioning/naming the disk. instead, they see mbr or gpt partition.
for example, when you go to install openbsd on a disk, it will add any foreign
partitions it finds to the disklabel starting at i, but that's simply so
openbsd itself can see and operate on those areas of the disk outside of the
bounds of gpt/mbr openbsd partition.
if you had two foreign partitions, they appear as i and j. but note, if you
create those partitions after openbsd, you have to add them manually to the
disklabel, which can be very difficult as there's no mechanism so sync this
information.
so yeah, generally want to avoid that.
there's nothing inherently special about the letter i (or any subsequent
letter through), i.e. they're not reserved exclusively for foreign partitions,
if none are found you can use them however you want.
the fstype is perhaps a better indicator here, but, also just a hint as to the purpose.
see also disklabel(8), openbsd faq - disk setup