Least Squares Adjustment for Level Runs
Without redundancy, least squares is a waste of time. You have to have observations to compare. That requires a loop, ties to other control or cutoff loops to analyze and adjust.
That's the bottom line. Nothing to analyze by least squares if you only have a single set of observations.
Personally, I add digital level files to my control networks, which also contain total station or GNSS observations, or both. It's easy to import and that way I have all my data in the same project.
But as others have said, with a calibrated (and recently pegged) digital level, unless you have a complex network of points, just adjusted by turns and leave it at that, it won't make much difference. Certainly not enough for 95% of survey work.
If you start from "Old A", run through five traverse points, then to "Old B".
Then run from "Old A", through "New C", to "Old B". you should have a complete loop... ?
I use LSA for my level runs. But I always have redundancies... i.e. run through GPS points as a check for busts. Ensure I run my levels in a loop. It gives me confidence that my level runs are accurate.
I think that you will find that most high order leveling procedures require that you run a loop between two known bench marks before you start running a new line. As others have said, unless you have a web of interconnected loops or are combining the level data with terrestrial or GNSS observations in a single adjustment then LSA will not give you any better information than a good old loop closure.
It will be a mute point is a couple of years since passive BMs will be obsolete. Beat the rush to clean you office and send my you LS15 now. 🤣
Closing on a second known benchmark is redundancy (minimal). By weighting the observations by either distance (I prefer) or number of turns, any misclosure will be distributed systematically.