Fix diagnostic behavior (maximum velocity, MAC velocity) on GPUs

[mbkuhn]

Summary

The unit test for the diagnostics on multiple levels was not passing on GPUs; the solution is to alter the way masking is used within the functions that find the max velocity.

Pull request type

• [x] Bugfix
• [ ] Feature
• [ ] Code style update (formatting, renaming)
• [ ] Refactoring (no functional changes, no api changes)
• [ ] Build related changes
• [ ] Documentation content changes
• [ ] Other (please describe):

Checklist

The following is included:

• [ ] new unit-test(s)
• [ ] new regression test(s)
• [ ] documentation for new capability

This PR was tested by running:

• the unit tests
  • [x] on GPU (gcc+cuda on Eagle)
  • [x] on CPU (apple-clang on Mac OSX)
• the regression tests
  • [ ] on GPU
  • [ ] on CPU

Issue Number: #1063