Improve accuracy of CoM integral

[iago-mendes]

Proposed changes

This PR changes the center-of-mass surface integrand so that the results are analytically the same, but numerically more accurate. It reduces the round-off error that exists in the nature of this calculation (see the comments in CenterOfMass.hpp for more details).

The plots below show the CoM results for different outer boundary radii $R$ for both an isotropic Schwarzschild test and an equal-mass non-spinning BBH initial data. The different columns show some of the integrand variations that we tested. Given the simplicity of using $(\psi^4-1)$ and its significant improvement in the BBH case, that's the version being used in this PR.

Upgrade instructions

Code review checklist

• The code is documented and the documentation renders correctly. Run
  make doc to generate the documentation locally into BUILD_DIR/docs/html.
  Then open index.html.
• The code follows the stylistic and code quality guidelines listed in the
  code review guide.
• The PR lists upgrade instructions and is labeled bugfix or
  new feature if appropriate.

Further comments

[nilsvu]

Doxygen is complaining

[iago-mendes]

Thanks! I had not noticed that. Doxygen is working now.