also, set "allocate_time" to early for particle datasets. it does seem
to have some negligible effect...

only test_io passes right now.

1 particle species seems to be correctly integrated at the moment.

kM is used in various computations, kMspec is the one relevant for
spectra.
And they are different for the smooth filter case...

in practice nothing changes, but the spectra will be shorter and the kM
eta calculation will be "correct".

stupid multifractality

also, fix max_estimate value in call to vorticity moments