The traditional North American split system heat pump (ignoring "dual fuel" systems for now) relies on a backup heat source to provide both makeup heat during low-temperature (below the system balance point) operation and, more importantly for this, outdoor unit defrost cycles. This backup heat is typically supplied by heat strips, and generously sized ones at that, considering that during defrost, the backup heat must carry the entire house heating load.
However, as documented in this Mitsubishi application note and this government research paper (as well as this research paper here), mini-split heat pumps stop or radically slow their indoor fan when defrosting to avoid filling the room with cold air, since they do not have a backup heat source to run during defrost cycles. This has the downside that they are not actively heating during defrost, but even without that, they are able to provide heating in most climactic conditions using this defrost strategy and sensor-based defrost cycle triggering without aid from a standby heat source.
Is there a reason that one could not use a couple extra wires in the outdoor unit cable and an inexpensive fan relay to retrofit a split system heat pump with the ability to perform a "fan stop" defrost cycle instead of turning the standby (backup, emergency) heat on? This seems like an obvious target for energy savings, given that emergency heat is far more power-hungry than heat-pump heat. Furthermore, even without any energy-efficiency benefits, this would likely allow the backup heat to be significantly downsized in climates where it's not heavily relied upon, as now it only has to be sized to make up the heat pump's capacity deficit at the heating design (99%) outdoor air temperature, not the entire house heating load. This can reduce electrical service sizes, and also position the system to be more easily handled by a backup generator.