If you built yourself a current limiter, that should stay pretty much dead nuts over the whole run as long as there's enough headroom above cell voltage for it to adapt to changing cell voltage.
The current limiter in power supplies has never been uber accurate at small currents. Its there as protection so that if you're powering something and it "runs away" or shorts out, you don't turn it or the supply into a room heater that might even set something on fire.
As cfnisbet says, our usage is not "normal" by any means. If you're limiting to say 1 amp, what does a fraction of a milliamp mean? Right? If you're limiting to 5-10ma, that drift is significant.
There is nothing wrong leaving it like this. If it stays OVER the target current by a little, just run the correct amount of time and add a bit more reducing agent and call it even. For things that go into the body, nothing is an exact science so as long as all the IS is reduced to Colloidal Silver, you're good. Whether you ingest 20PPM or 22-25PPM, it matters little though if you're processing cold, that could be an issue given you really don't want to go up over 21PPM or so as above this it will start precipitating out silver oxide which isn't good. If you're processing hot and adding the reducer at the beginning of the run, you're fine - just add a smidge more to compensate for any misgivings. I usually do this anyway. A little too much reducer does nothing bad. Its sugar and a TINY amount of it to boot (unless you're using cinnamon tincture but the extra just makes it taste a bit more cinnamony -wink).