However, playing with them and the probe, I’m not sure if it makes sense. If I push the probe all the way in, it will obstruct the flow, plus only the tip will be surrounded by fresh water:
I use a T connector (actually a pushfit connector). I protrude the probe only just into the water path, which I feel gives adequate contact without unduly restricting the flow. The metal casing has, I believe, better thermal conductivity than the water.
I actually have two sensors, one before and one after. I feel that the one before is required to ensure I’m providing sufficiently cold water and the one after shows, from the relative temperature, whether there is adequate flow.
I do have a flow sensor too, since the temperatures would not reflect the electrode temperatures if flow stopped.
This is overkill, of course. But who’d want to oversimplify it ?