I'm not an electronics wizzard but I have wired in a few things. But I discussed the quote from the TSRM with an electronics engineer (who btw has build his own ECU for a 7mGTE from scratch) and he told me it would be possible for an ECU to distinguish an unplugged sensor from a broken sensor. But on the JZA80 doing so through VCC would be difficult since VCC goes to both TPSs and to the MAP-sensor.
But in general an ECU could test either current through VCC-E2 or connection through IDL-E2. This combined with short or open i VTA-E2 could be interpreted by the ECU as sensor present but broken. No current through VCC-E2 or no connection through EDL-E2 and open in VTA-E2 would then be interpreted as sensor unplugged.
But I fail to see how distinguishing between the two situations (plugged or unplugged sensor) will be of any use since both a broken and an unplugged sensor will need fixing. And in the situation with testing for present sensor through connection through IDL-E2 an unadjusted broken sensor will set no codes.
So I think that the quote is wrong too.