In Facter 4.* I'm missing the ipaddress_em0 fact. Is this a regression in Facter 4? It was regenerated in d3489c4 but there it's also missing.
Indeed, these facts are not generated on Facter 4 and have probably been missing for quite some time (since a FreeBSD specific networking structured fact has been introduced probably).