Behavior of combining Loopback timing reference and Offset

Does using a Loopback timing reference automatically compensate for the timing differences observed bout output and ref in a measurement?

If so, what is the effect of having both the loopback and the offset?



I got the above by using a loopback reference, and then using "Estimate IR" to update the standing offset for future measurements. Do the authorities fight with each other in establishing t=0, or does the loopback reference always win?

I see. So even if you use a loopback timing reference, you pretty much always have to use "Estimate IR delay" on the result, unless you are 100% sure nothing moved since the last time the offset was updated?

But if "Estimate IR delay" is required, do I actually get a benefit from using a loopback reference at all? Won't I get the same accuracy if I just use "set t=0 on IR peak"?