This link has the arxiv report too.
https://ui.adsabs.harvard.edu/abs/2020arXiv200600645C/abstract The abstract states ""We numerically explore the possibility that the large orbital inclination of the martian satellite Deimos originated in an orbital resonance with an ancient inner satellite of Mars more massive than Phobos. We find that Deimos's inclination can be reliably generated by outward evolution of a martian satellite that is about 20 times more massive than Phobos through the 3:1 mean-motion resonance with Deimos at 3.3 Mars radii. This outward migration, in the opposite direction from tidal evolution within the synchronous radius, requires interaction with a past massive ring of Mars. Our results therefore strongly support the cyclic martian ring-satellite hypothesis of Hesselbrock and Minton (2017). Our findings, combined with the model of Hesselbrock and Minton (2017), suggest that the age of the surface of Deimos is about 3.5-4 Gyr, and require Phobos to be significantly younger."
The surface age of Deimos is critical here and reconciling the age differences between Phobos and Deimos. I am glad to see *we numerically explore the possibility* in reports like this. The space.com article stated "Deimos is billions of years old, but Phobos is as young as 200 million years old — meaning it formed when dinosaurs roamed the Earth."