fix an unintuitive behaviour in ProfilePlot

[neutrinoceros]

PR Summary

Because of the way it's currently implemented, ProfilePlot produces unintuitive results when passed x_log=None, which is effectively, equivalent has passing x_log=True despite the fact that bool(None)==False.

with this change, the following produce the same result

import yt
ds = yt.testing.fake_random_ds(128)
yt.ProfilePlot(ds, 'y', ['density', "velocity_magnitude"], n_bins=500, x_log=None)
yt.ProfilePlot(ds, 'y', ['density', "velocity_magnitude"], n_bins=500, x_log=0)
yt.ProfilePlot(ds, 'y', ['density', "velocity_magnitude"], n_bins=500, x_log=False)

This issue was reported on slack by Jack Jenkins.
I'm opening this as a draft since there may be other appropriate changes to ProfilePlot.init() that could go with it.

[munkm]

I have a few comments! But this is a good idea and I agree the way that we're currently that using None for these log fields is confusing. We need to be really clear in the release notes that this may cause different behavior in scripts of our downstream users.