Why is there a check valve on the raw water input to watermaker?

[accomplice]

Our SeaRecovery Aquamatic recently stopped working. It failed reporting that the output of the booster pump was low. In troubleshooting the problem, I discovered that a one-way/check valve on the raw water input was impeding flow, reducing it from ~20gpm to less than 1gpm. Removal of the check valve solved the problem and the water maker works fine. There are two things I do not understand, though. First, why is the check valve there at all?

A little background on the installation. Raw water flows from large through-hull to a sea chest. The sea chest feeds several consumers, including two heads, two A/C units and the watermaker. The feed to the watermaker is configured through a 3-way valve to allow the source to come from the seachest or from a bucket. The hose from the sea chest is (was) then sent through a check-valve, then to a T. One side of the T is a sensor for raw water termperature and salinity. The other side of the T goes to the booster pump.

The check valve was located approximately 10" below the waterline. The pressure from this drop was insufficient to open the check valve more than a trickle. If the check valve was lowered another foot or so, with nothing connected to the output, it had full flow. In other words, it would seem that the check valve was working, but required greater pressure than the drop below waterline provided. I'm not sure how this could've changed on my recent passage (it worked on day 2 of the passage, but not day 4). This leads me to my second question -- why would this configuration have worked in the past, and then stop working?

Lastly, is there any reason not to (permanently) remove this check valve?

Thanks for your help....