vertical manometer

[Tony]

John makes a very good point about a slightly larger bore manometer tube draining back faster. It will also have fewer problems with trapped air bubbles, and be easier to see.

For velocity probing, minimum water volume in the manometer will always respond faster (water is heavy!)
But for test pressure measurement, a really fast response is not such an important requirement.
My choice would be a single vertical glass tube feeding from a reservoir of relatively large surface area.

Overpressure accidents are not that uncommon, and it is very easy to blow the water right out of the top of the test pressure manometer. An upper overflow tank that drains back freely works well.