I don't think there is an "agreed distance".
It's important to be aware that a mechanical vario is in effect a controlled leak.
So when air pressure goes up, the mech vario "steals" a little bit of that pressure.
When it goes down, the mech vario "puts back" a little bit of that pressure.
It does this ALWAYS. The more sensitive the mech vario is, and the larger it's capacity flask, the more this effect shows up.
So in general, follow these rules:
1) If at all possible, don't put anything else on the same tube as a mechanical vario
2) If #1 is not possible, make the split as far from the mechanical vario as possible, to reduce this effect.
In my personal opinion, the length won't have much effect on reducing the error introduced by the mech vario.
In theory, it should due to resistance of air moving through the tube, but as we know, that resistance is minimal.
A 3rd option is to put a restrictor on the mech Vario, so it behaves more like a smoothed average, as that would reduce the interference.
Assuming that the mech vario is a backup to the primary electronic vario, consider not having the mech vario be a TE vario at all.
Why mess up your primary vario, to have a good backup that you rarely use?
Anyway, hope this helps.