Fitness fixes

[dfalster]

Resolve some outstanding issues new mutant fitness method introduced via #362

Specifically,

• caching of resident results enabled when using fixed steps (in inst/include/plant/ode_solver.h)
• offspring production was not being recorded properly in
• the fitness_landscape function was producing values that were different to running mutants directly. This turned out to be because the knowledge of node timing was lost with use of remove_residents. this only became evident once running with higher densities for residents, which we didn't previously test.
• add a mutant_parameters method
• more extensive tests

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

:exclamation: No coverage uploaded for pull request base (develop@51b8f99). Click here to learn what that means.

:exclamation: Current head fb35498 differs from pull request most recent head c1ea9a2. Consider uploading reports for the commit c1ea9a2 to get more accurate results

Additional details and impacted files

```diff @@ Coverage Diff @@ ## develop #379 +/- ## ========================================== Coverage ? 82.93% ========================================== Files ? 99 Lines ? 9038 Branches ? 0 ========================================== Hits ? 7496 Misses ? 1542 Partials ? 0 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.