EM, yes in the absence of current.  If BSP/SOG% > 100, your speedo is over-reading at least for that combination of BSP and Heel.
I always look at dBSPSOG as a function of Heel and BSP in Race Replay also, since the issue is often there rather than in the speedo calibration itself.

nav, good question, I don't know.  When you look in Race Replay Team version Config file (I have Team), there are 4 variables tabs - Known Variables, Other Log Variables, Calculated Variables and Manual Entry Variables.  I have three or four entries in Calculated Variables, one of which is BSP-SOG.  This enables you to produce XY graphs in Race Replay.
In PRO version, you could do the same thing by producing a normed Phases report and doing the calculation in Excel ...... but it's easier not to.  I do this anyway for analysis of angle/sail performance, I find that easier in Excel.
PM me and I can send you a screenshot of what I see if this doesn't make sense, and I guess we could probably look at the KND website .

OK, but by and large a known quantity in that part of the world.  Three ways:
1.  Going around marks in a known current area and time, do you see the same dBSPSOG difference both ways - not sure if I have expressed this well, but you know.  Do this graphically as in #3;
2. Does average dBSPSOG on any leg in known currents areas match what you expect?
3. Draw a part of the logfile - either in Expedition or RR Chart Focus with current barbs switched on - does the result look like you would expect, given the known current at that time?  Generally better to look at a portion of the day/race i.e. known current for a period, rather than he whole day/race at once, which could well muddy the result.
And as I well know, if your compass is out this analysis will cause confusion ... so make sure that's OK first.