So you measure the resistance and adjust according, every tech worth a crap knows this. Of course, that is assuming the meter is worth a crap.In that case, I'd agree with 2 x 1.2k in parallel, because that gains the greatest benefit of resistors in parallel decreasing error due to tolerance, and also good for stockholding!
This website
http://sim.okawa-denshi.jp/en/rercal.php
will calculate optimal resistors in parallel, potential dividers and loads of other stuff.