When small (about 10 meters or 33 feet) near-Earth asteroid (NEA) 1991 VG was discovered in 1991, it was the first one of its kind and it attracted plenty of attention. Its preliminary orbit was the most Earth-like known at the time. Lacking a proper context, the new discovery was not easy to interpret. It could be a natural object, the first representative of a new type of asteroids, or it could be space junk, hardware left behind by the Americans or the Russians.
If natural, it may have had an origin in the main asteroid belt, the Moon (debris ejected during a relatively recent impact) or perhaps near-Earth space via collision or fragmentation events. If artificial and given its likely size, it could have been one of the stages of a heavy-lift rocket, a large satellite (functional or not, well-known or part of some secret project), or even something literally out of this world, an alien spaceship.
As the orbit determination of 1991 VG got improved, it became clear that some of its dynamical properties did not match those of the few known (back in 1991-1992) NEAs. Its orbit was similar to that of our planet, but in such a way that the object was able to return more or less periodically (every few decades) to experience two or three relatively close fly-bys with the Earth (and the Moon) in as many years. It did not look particularly dangerous, the probability of colliding with our planet was insignificant. However and from the point of view of the Earth, it engaged in some sort of Moon-like behavior for a few months. This was truly unexpected and resembled what an active space probe is expected to do when studying a target body (planet, asteroid or comet, for example). From a certain point of view, 1991 VG became a (sort of) second moon of our planet in 1992, but this was a very short-term engagement.
An object so small was soon out of reach of ground-based telescopes. Having been discovered on November 6, 1991, the last observation was recorded on April 27, 1992. The object has remained unobserved until recently, having been recovered from Cerro Paranal in Chile on May 30, 2017. The new images came from one of the largest telescopes in operation today, the Very Large Telescope (8.2 meters). Although not observed for over 25 years, the object continued fuelling the imagination of the public and experts alike. With the new data and a  much larger population of NEAs to compare with, it was time to confirm or reject some of the early speculation about this object. The new observations have improved the orbit determination significantly. It is now clear that its path is not so Earth-like as initially thought. There are plenty of known NEAs which follow much more Earth-like trajectories; they are designated by some as the Arjunas.
More information is provided here: Geometric characterization of the Arjuna orbital domain.
The Earth has even a few quasi-satellites or asteroids which have orbital periods nearly equal to that of our planet and that never stray too far from its neighborhood (although they are transient companions).
More information on this subject is provided here: Asteroid (469219) 2016 HO3, the smallest and closest Earth quasi-satellite.
The orbital evolution of 1991 VG as computed using the latest data confirms several early findings: the recurrent fly-bys, the virtually zero risk of colliding with our planet or the Moon, and the fact that 1991 VG engaged in satellite-like behavior with the Earth back in 1992. Other dynamical details are similar to those found for other NEAs moving in orbits similar to that of the Earth.
See, for example, A trio of horseshoes: past, present and future dynamical evolution of Earth co-orbital asteroids 2015 XX169, 2015 YA and 2015 YQ1.
The new data show that no known space mission matches the orbital and timing properties of 1991 VG. Therefore, it is very probably not space junk.
As all the properties of 1991 VG derived from the data (new and old) can be easily explained within the context of the known NEAs following paths similar to that of our planet, there is no need to attribute an artificial origin (alien or Earthly) to this small but dynamically interesting object. These are some of the findings discussed in our work on the dynamical evolution of 1991 VG: Dynamical evolution of near-Earth asteroid 1991 VG.
This work was led by Carlos de la Fuente Marcos from the Universidad Complutense de Madrid.