SOG vs SOW for RADAR /ARPA

I know this is a controversial topic , but would like to know your practical view points .

My view is as below ,

SOW(Speed Over Water) should not be used in RADAR at all as none of these instruments are showing any consistent reading . 

The main argument given for favoring SOW is the aspect may not be detected correctly if SOG is used , yes this is true but when aspect is a requirement in taking avoiding action then it means vessels are in sight of one another (Section II - Conduct of vessels in sight of one another (Rules 11-18))

If the vessels are not in sight of one another (Poor visibility ) then finding our aspect also has no value .

Many times if you have used SOW then you will find the vectors are never steady as the EM Log or Doppler log has huge variations in s short span of time . In case of Doppler log it may even change to SOG when the depth is low .The wind drift factor is not even considered in this assumption , the notion that only current will affect is not correct and in low depth areas current may affect differently (places like Singapore straights or coast of Japan etc ) . The accepted view that SOW water to be only used for collision avoidance is something which requires to be questioned , but sadly in this industry we tend to have some holy things and SOW is one such thing . 

Your equipment error and your drift is what you are applying to another vessel and getting the aspect , how correct it can be if on a heavy beam wind one PCC and a fully loaded VLCC meets and gets the aspect by applying SOW ?.

I would like to know if you are sailing now or when you were sailing what were you using and how reliable was your EM or Doppler log , was the vectors steady?

In this age of AIS you have the Gyro heading of the other vessel readily available so the aspect finding by your ambiguous sea speed is not even required , this is in case if you are really interested in finding aspect electronically .