My electronic strobe tuner allows me to tune reeds nominally to the nearest 0.1 cent, and that's what I do - trying as far as possible to keep all other factors as controlled and as constant as possible, e.g. especially bellows pressure, reed block stability/rigidity and temperature.
I've successfully tuned around a dozen or so instruments now. I know that's nowhere near as many as Theo or Rees or any of the other tuners around, but I am learning all the time, getting a feel for my tools, individual reed types/makes, and my listening skill. I am perfectly aware that there is an 'uncertainty principle' operating. As mentioned earlier, after tuning a reed carefully, I know that it will almost certainly drift slightly away from pitch after a few hours, for whatever reason. A second check and tweak of the tuning usually stabilises things much better and again as mentioned earlier, if on checking I find the reed to be within 0.4 cents, I leave it alone.
But since I am able to initially tune to the nearest 0.1 cent, I see no justification in doing otherwise. If the uncertainty principle means that the tuning is going to drift by say ± 1 cent or even more in the hours and days of playing immediately following tuning, at least I will have started out from an accurate baseline, even if the 'error bars' either side are relatively large.
The drifting and the magnitude of the error bars don't really matter too much when the requirements are for a fair amount of tremolo, but it does become far more critical for the upper notes of an instrument tuned with a very light tremolo: e.g. for the note B5, an error of +1.25 cents for the M+ reed and -1.25 cents for the M- reed is the difference between a 1 Hz tremolo and a 2 Hz tremolo, which is easily noticeable.
Another example is the 'bandoneon type' tuning of L and M reeds in octave unisons. I have found that even a -0.5 cent error on the L reed can cause a noticeable difference; there can be an audible roughness to the start of the note as the lower reed is slower to achieve its stable pitch. Eventually the two reeds will pull each other into a compromise unison, but the delay can be minimised by such careful and accurate tuning. As ever, ultimately the ear is the final judge.