Make observables run in parallel

[reinaual]

Fixes #3154

Description of changes:

• make observables run in parallel
  • only send the extracted data via MPI (the original implementation sent the entire particle data structure)
  • remove a major performance bottleneck in parallel simulations
• remove 8 callbacks from the deprecated MpiCallbacks framework

[jngrad]

Here is an updated version of the plot #3154 (comment) which uses the new implementation of the observables, using a TimeSeries of the particle positions:

The benchmark writes to an SSD, either raw unsorted particle positions for MPI-IO every 20 time steps, or sorted particle positions for observables written once at the very end (sampled every 20 time steps with an auto-accumulator).

Attachments: benchmark.py, plot.py