Years from now, new benchmark revisions will come along and change the whole picture again -- which begs the question of how important these monthly payroll changes are. I think it's irresponsible to make too much of these reports.
Benchmark revisions could completely recast underlying data, in which case January forecasts won't be useful.
That kind of decline is a perfect fit with what we've seen in the household survey's measure of employment, ... if not in February, then in future benchmark revisions.