Got sick of the "he said/she said" arguments here and went to the horses mouth: OVID Medline Medical journal database.
With the following search:

I searched for dentifrices, baking soda, and dental plaque.
Combined them with AND
Limited results to articles published in this century.
In total I had 8 articles that were relevant:

Reading through them...
Ughhh. So much bad science.
Like the one funded by a company that produces a certain toothpaste, without further clarification of how much input the company had. There is also no mention of how the groups were randomised, no standard deviation, and no mention if any people dropped out.
It's so bad that I would not rely on the article any more than saying that "some toothpastes may decrease plaque amounts when compared to no brushing".
I think I'll just stick with my standard fluoride toothpaste.
Men wanted for hazardous journey. Low wages, bitter cold, long hours of complete darkness. Safe return doubtful.