I might chime in with a few qestions myself

, others may be wondering the same thing/s...
A bit of preamble first tho....
If when we are measuring air flow in a duct using a pitot tube we are measuring two pressures.... velocity pressure (Pv) and static pressure (Ps)
Pv is as a result of the velocity of the air travelling in the pipe and is measured with the inlet to the pitot tube pointing direct into the air stream
Ps is from the restriction to air movement created by the fan...small pipe, lots of bends, long run, blocked pipe etc. etc. and is measured at 90° to air flow
The question/s are...
Is there a difference of Pv values if we measure on the suction side of the fan when compared to the exhaust side?
Is the Pv higher by the amount of Ps if we measured on the exhaust side?
and
Is Pv lower by Ps on the suction side?
My gut feeling is yes on both accounts but I could be wrong
Would I get acceptable Pv results if I simply pointed a piece of 1/4' or such sized steel/brass tube into the airstream?
or
Is the entrance into the Pv port something special fit for the purpose? Sized for the proposed expected air speed maybe?
The one I have used a long time ago did have a small hole at the inlet, maybe 1 to 1.5mm, this possibly dampens the fluctuations to give a steady reading in the manometer but Is there more to this? It was also bullet nose shaped at the inlet.
I read on Bill's site that the distances from the bend, Ps port and the Pv inlet in relation to each other (for a combined tube) gives the pipe a C value that is included in the calcs but if we are measuring with separate tubes as in BobL's pic above do we need to consider the C value?
I wonder if a few pics of the end of your pitot tube might be possible Bob?
Intergrating...mmmm it's been far too long since I have done any of that...
What was the reasoning behind needing to do that? Bob/Graeme has said that from an accuracy standpoint that a closer to a real value is obtained but in my head it's only mud

mmm thinking thinking maybe something to do with straightening the curve to give a linear result aarrghhh to long ago I can't remember

That's probably enough for the mo.....
Pete