ParticleContainer::RedistributeCPU for Pure SoA

[ax3l]

For pure SoA particle layouts, we see on CPU that ParticleContainer::RedistributeCPU takes 8-9% longer than the legacy layout. BLAST-WarpX/warpx#4653

I think we can probably avoid this by optimizing the pack/unpack routines for the new layout a bit more:

• check and avoid if we still use the legacy aos layout for some of the communication - ensure we perform nice and long array copies
• use the new id validate/invalidate routines from Add ParticleIDWrapper::make_invalid() #3735

x-ref: #2878 #3299

[asalmgren]

Closing this issue due to lack of activity. Please re-open it if you feel there is is still an issue!