Well the question on the Roche limits as i say is two fold, one, the two models are quite fuzzy as you already say, they are completely about the formation of large objects close to each other and a point where tidal stress becomes an important consideration. I have no major dispute at all with the treatment and the rough pick of densities, but treatment of the concept is still, in my opinion somewhat vague.
The Roche lines and limit represent the point at which accretion of material onto a seed body, is surpassed, it limits the size of an object on the sole basis that gravitational attraction to the primary body is higher than for two bodies next to each other...and say, a rock on the surface of a body is to its major mass.
There are also two ways in which you may model the rings, and its something i want to check out... which is, rings have a none zero thickness, we know the inner and outer radius and the mass. If modelled as a solid disk, of material of that density, it gives us another option of how the system behaves as though it was rigid. Now thats clearly not an amazing treatment, but it is also relevant, the gravitational field experienced by an outer ring is different to an inner ring.
Also the band structures we observe in rings is another indication our model is rather simplistic and using it as some kind of yes/no absolute is also stretching the meaning of the model. This banding is due to tidal resonance and the presence of shepherds largely.
The formation of a ring also plays a very important role and the presence of other planetary bodies, a gas giant, on its own, may have a ring that is far older than one contained with lots of moons. A planet on its own, its rings will survive as long as it takes to accrete material back onto the primary, and or create a moonlet from the material, this likely wont form right on the roche limit, but in excess of it.
My point about limited data, is true in every meaningful manner, we do know the physics very well, but what we don't know for sure is the absolute history of events and the true ages of these systems. Physics models are quite accurate, but they always include assumptions. Similar to the Roche limit being treated as solid or liquid... while the liquid limit is probably closer to the truth it really still is an assumption and knows nothing of material strength, sheer and elastic limits for the materials in question. Why? Well because its more trouble than its worth and would probably just give us a stable extension outside the roche limit by .... some factor. But its not worth the worry because ball park the Roche treatment appears to work.
It is sort of the difference between trying to examine exactly what is going on, as best you can, or, going with something that seems to fit and stopping there. We don't expect physics to change, but, that is assuming we modelled all the physics correctly in the first place...which... we really have not
And also of course, there are many misconceptions about how it applies to solid objects vs spherical bodies in hydrostatic equilibrium, etc, but getting a bit off topic.
