Happy to hear your zip tie fix is holding up. Sometimes simple is the best!
I thought I would update this for others to refer to during their design/build process.
I decided to implement water cooling by ram feed only (no pump)
I 3D printed a U shaped fitting that acts as a water return from one channel to other.
I used the original push in water fittings on the 75200 to connect the water tubes (quite a few sessions now and no leaks) so in my opinion no need to change those from original.
I used 8mm OD - 6mm ID (thick walled) plastic tubing to connect to the inlet and outlet channels of the vesc
I did a test today to determine board speed required to get flow. I added a short extension to the outlet side and aimed it vertically upwards to create a vertical lift of approximately 100mm from the intake to simulate anyone using a tall mast (normally my outlet tube is pointed down towards the water at the level of or below the outlet of the vesc to minimize the vertical lift of the water). At approx 5 km/h I started to get a trickle of water and as speed increased the flow gets stronger.
I donβt think it takes much water going through the vesc to keep it cool and my logs have never shown any increase in temp (within 1-2 deg C) and that is confirmed by touching the vesc at the end of a session where there has been substantial sustained current flow as Iβm in the learning stage so have an above average number of starts in the session.