Hi James,
I have a similar issue. I have not re-done delineation - I rather suspect that this is due to the fact that there are "lakes" in the model set up (as also seems to be the case in the post above). While I cannot check the Toolbox code as this is not yet released, it seems to me that SWAT+ Toolbox looks up the channel numbers in the "SWAT+ unit" column of channel_sdmorph_day.txt file. If this is the case, this will not work for SWAT+ set ups where lakes/reservoirs are included. When including lakes, QSWAT+ will remove some channels during delineation (i.e. those inside a lake), and therefore the number in "SWAT+ unit" is no longer equivalent to the "gis_id" or "name" columns in channel_sdmorph_day.txt (which represent the channel numbers in the rivs1.shp file).
I guess a work-around would be to first identify the "SWAT+ unit" for the corresponding channel number (in channel_sdmorph_day.txt), and use this when defining the "Object Number" while loading the flow observations into SWAT+ Toolbox?
Cheers,
Dennis
PS: to illustrate the point, I attach an extract of the channel_sdmorph_day.txt output from the set up I am working on, which includes several lakes.