The Redarc is connected directly to the load batt (and nothing else other than charge source, and possibly solar - a secondary charge source). Every 100 seconds it ceases operations, takes a measurement of the load bat, and carries on till it hits the selected charge profile limits.^ OK, not sure if I got that ("knows the status of charge") right, but sounds like the weak point in the approach you describe is that the charger has to know the level of charge of the battery at all times.
Short of measuring directly the voltage there, it'd have to do some historical calculations based on some presumably known initial condition. IOW, keeping track of how much energy has flown into the battery knowing how charged it was before.
It doesn't track total energy, just how far off the load batt is from the hi/lo values in the charge profile it's using. I use a monitoring device to track total energy, tho I haven't found the data to be useful other than making me feel like a wasteful human. lol
But the monitoring does contribute to my explanations here as I have about a years worth of base lining as a result, watching the Redarcs draw on its charge source and its output to the load source across all 3 charge profiles at differing charge states between the front and the rear batts.
For example, my AGM batts have their charge/low voltage levels on a sticker. The Redarc profile B matches those specs exactly. So if I leave the Redarc to use profile B (I can switch between profiles from the dash, and do sometimes), and if the truck runs long enough, the RedArc will not leave bulk mode until it sees the charge portion (bulk) of profile B has been reached.That would be quite an iffy proposition, I think, given particular unknown battery specs and condition, inefficiencies and the like. I suppose you could charge something like that but to expect it would be charged optimally that way seems like a long shot. Then again, I'm probably missing something. Interesting topic.
Certainly can't speak for all setups, but I can say there's been no iffyness in this. When the RedArc hits a mode other than bulk, the load batt(s) have always been fully charged (or near enough ~29.6v across 2 - 12v batts rated to 14.8v/ea max). They settle down a bit moving from bulk to absorbtion, but that's SOP and part of getting the last little bit in the batt.
Capacity decay in all its manifestations can be accounted for somewhat, but it's still a net sum loss as in any other situation.
There's a reason the RedArc is ~$500 and most of it's competition is not.
C-Tek comes to mind in price parity. Not too familiar with it, tho I imagine it should work the in same way.