The beginning of the abstract: "Numerical modeling has long suggested that gravitationally-bound (or so-called rubble-pile) near-Earth asteroids (NEAs) can be destroyed by tidal forces during close and slow encounters with terrestrial planets. However, tidal disruptions of NEAs have never been directly observed nor have they been directly attributed to any families of NEAs. Here we show population-level evidence for the tidal disruption of NEAs during close encounters with the Earth and Venus."
Sure, that's fine, but that's not something that really needs evidence.
General relativity is something that needs evidence. It EXTRAPOLATES new knowledge from our existing knowledge.
Loosely bound asteroids, on the other hand, are things that are well within our knowledge.
I don't need an evidence that throwing an astronaut off an airlock is going to kill them. We've never tried doing this but that doesn't stop us from being able to pretty firmly say that they are going to be dead pretty quickly.
For asteroids, we know that lots of them are just piles of rock unbound in any other way than gravitationally. We have actually shot them with things to verify it. We have well tested formulas that are not extrapolations that tell us exactly what is going to happen when we bring together a loose pile of rock and start spinning it. We also have formulas that can tell us exactly what kind of forces are going to be acting on a loose pile of rock of certain mass and diameter when it is shot through a gravitational field.
There is very little uncertainty about what is going to happen. Just because we have never saw it happen does not mean we need any further evidence for this. Doesn't mean there is anything wrong about preparing evidence. It just means it is not very interesting.
