Round after prime mapping not in prime mapping

[frostburn]

https://sw3.lumipakkanen.com/scale/x5igOPjMt

galaxyspoob — Today at 12:37 PM

In SW3, when CTE tempering is applied where the equave is split, the fraction of the equave (the half-octave in the below example) is sometimes off by 1 minimum unit of interval size, no matter the setting of cents digits after the decimal point.

E.g. when https://sw3.lumipakkanen.com/scale/x5ijC5mUP is tempered by {1716/1715 2080/2079}, 99/70 turns into 599.999, 599.9999, or 600.00001c for the 4-, 5-, and 6-digit settings.
(The JI scale was made by writing the first 5 notes, then Repeat, then the manual replacement of 9801/4900 by 2/1.)
(Raw version): https://sw3.lumipakkanen.com/scale/x5igaO9ML
Ah, so that occurs because PrimeMapping is rounded.