No use worrying about could haves and should haves. Factory chambered barrels are hit and miss, some never shoot properly, some shoot properly nearly immediately. Just start where you are and move ahead and log your data. By 100 rounds the barrel should be 'broken-in' weather you followed a regimented approach or not. Just verify where it is in the process, maybe it is already good to go! No need to waste a bunch of barrel life by starting over, the good news is that you have some fire-formed brass at least!!
Before you go any further, use a quality copper removing solvent and make sure to get every last trace of copper out. Once you are back to that state fire five rounds over a 15 minute span of time over your chronograph. A magnetospeed or a Labradar are your best bet as they produce accurate velocity data. If all you have access to is an optical chronograph, I'd still use it. However, understand those aren't precise instruments, some data is better than no data, and log the velocity.
The 28NOS wasn't ever meant to fire strings, so take your time between shots!! De-copper the barrel again and repeat. You are looking for velocity numbers that are pretty close the first set of numbers. If they are then you can be certain you are ok to start load development.
If they are not close, don't freak out, just de-copper it again after five rounds and start a seating depth testing with minimum charges and keep logging the data. One depth will print better than the others. No need to shoot anything loaded hotter, as that just puts more wear on the tube that you don't need to put on it. As a barrel settles in it will speed up, some more than others, then stabilize once it reaches its optimal condition.
Once I'm certain a barrel has reached its optimal condition I don't clean it again until I have a legitimate reason to clean it. Such as accuracy begins to degrade or SD isn't holding up with the same lot of ammunition.