small inconsistency in accounting for capac residual?

[gmao-rreichle]

@rdkoster:
Near the end of subroutine catchment(), we compute (total) RUNOFF as the sum of surface runoff and baseflow:

GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatch_GridComp/catchment.F90

Line 1271 in d21d478

RUNOFF(N) = RUNSRF(N)+BFLOW(N)

Immediately thereafter, we check for negligible amounts of CAPAC (<1e-10) and convert them into (total) RUNOFF:

GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatch_GridComp/catchment.F90

Line 1273 in d21d478

RUNOFF(N) = RUNOFF(N)+CAPAC(N)/DTSTEP

Don't we also need to add CAPAC(N)/DTSTEP to RUNSRF (or, alternatively, to BFLOW) so as to maintain consistency with (total) RUNOFF?
The water amounts involved are probably so small that in practice nobody ever notices. But we go to the trouble of accounting for them in RUNOFF, so maybe it's worth to also adjust the runoff components for consistency.
Please advise, thanks

PS: Here's the full block of code for reference (lines 1271-1275 in today's version of the "develop" branch)

        RUNOFF(N) = RUNSRF(N)+BFLOW(N)
        IF(CAPAC(N).LT.1.E-10) THEN
           RUNOFF(N) = RUNOFF(N)+CAPAC(N)/DTSTEP
           CAPAC(N) = 0.0
           endif

cc: @biljanaorescanin