As a practical matter (long time developer speaking), this doesn’t have to be true. It certainly makes sense that transition detection, assuming it happens in the device firmware, can only detect transitions related to CT’s hooked up to that device. As long as a multi-sense installation does not have CT legs across “boundaries” so that detection would require combining data from two sense-units, this should be possible to (make it) work. Even if detection happens on the mothership.
It seems the basic requirement is the notion of tagging all data with a sense unit id (which I am sure already happens in some form). As long as multiple units are registered to a single “owner” it should also be possible to mmodify all data queries currently used to not select based on a single device id, but on the union of all “owned” device ids. That, then would allow a unified look at most of the data.
The issue is slightly different for scenarios where multiple sense units are in use, but on different properties, because in that scenario, you would want an easy way to select which property one is looking at, but once that is done, you would not want data merged.
A final hurdle could be for what is called aggregate data, such as total consumption, because those could be looked at on a per sense-unit basis (assumign it is even configured to measure any kind of consumption), or across all units on a single property. That would definitely require some additional work. I bet, though that most users that have asked about multi-units would be happy even if the aggregate totals where not there yet.
Homeassistant (I am a user) would seem a way around some of these issues, but in practice it will not be so easy. This is due to the fact that all current “integrations”, the means by which one access data from sense for example, only allow configuration for a single user/account at any one time. Since sense has a one-device, one-user, policy, this effectively makes unified management, data combination/aggregation/presentation, quite hard unless one goes to greatly expand functionality of existing integrations. Additionaly, I am sure there are many users with multi-sense desires that are not capable, or willing, or have the time, to go program homeassistant.
So while I do not accept the premise that these things are not possible, or even technically difficult (with a few exceptions), I do also not expect any of this to be done. The reason should be clear to most by now. Sense is entirely NOT focused on serving the customer base of the sense unit. Its energy seems to go into other business, probably bigger business with utilities. The current product has not really received any substantial functionality enhancements in a couple of years now, and even responses to alledged bugs are slow, and sometimes, non existant. James is doing his best, but he is only the communicator and does not determine company policy.