Looking at the actual numbers it seems like it's all within the margin of error and the imperfections of DAD. Sometimes Foobar scored better at some test, sometimes JPLAY.
Meaning the measurements detect nothing but statistical noise?
No ABX listening test seems to be required after that result.
From my understanding of RMAA test, yes, it would seem so.
Here are the results for those who don't want to bother downloading the files.
Foobar (second run at 44.1 aka "[MME] 2496 WAV+in Foobar-2.htm"):
Frequency response (from 40 Hz to 15 kHz), dB +0.02, -0.07 Excellent
Noise level, dB (A) -100.3 Excellent
Dynamic range, dB (A) 100.8 Excellent
THD, % 0.0007 Excellent
THD + Noise, dB (A) -89.8 Good
IMD + Noise, % 0.0038 Excellent
Stereo crosstalk, dB -99.6 Excellent
IMD at 10 kHz, % 0.0036 Excellent
JPLAY (second run at 44.1 aka "[MME] 2496 WAV+in JPLAY Foobar-2.htm"):
Frequency response (from 40 Hz to 15 kHz), dB +0.02, -0.07 Excellent
Noise level, dB (A) -100.3 Excellent
Dynamic range, dB (A) 100.7 Excellent
THD, % 0.0007 Excellent
THD + Noise, dB (A) -89.8 Good
IMD + Noise, % 0.0038 Excellent
Stereo crosstalk, dB -99.5 Excellent
IMD at 10 kHz, % 0.0036 Excellent
(Sorry, I don't know how to align those columns for a better view.)
These are the two closest results. The biggest difference was in the first run, with a 1.1dB difference in Stereo crosstalk (-99.7 vs -96..
So yeah, I'm pretty sure none of these differences are audible.