It has been a while, and it is fair to look back and see how I did:

Quote Originally Posted by jerry
There are a few interesting things coming down in the next 18 months or so: Direct observations by NEAR in the asteroid belt; Stardust pays a visit to the Deep Impact target, Tempel 2. Messenger goes into orbit about Mercury and Hayabusa returns. Lots of excitement.

Tempel II: Mission scientists expect to see a gaping hole carved by the Deep Impact probe. I expect an unobservable-to-barely observable surface-wound feature. The difference depends upon whether Tempel II is a dense object, or an object with a density of less than 0.2g/cc, as Deep Impact scientists calculated. I think that is absurd.

Messenger: With a successful orbit, Messenger will be able to measure the gravity fields of Mercury. We won't know right away, but over time, a painful discrepancy will emerge: the high regions will map out as being unphysically underdense, while the low regions will map extremely over-dense. Expect a delay of at least six years in the release of the data: It is hard to release data that is not understood. (The oddball changing color of Pluto was first observed eight years before this was published.)

Near: Just as engineers had a devil of a time figuring out how near Hayabusa was to Itokawa; Near is going to find navigation very close to asteroids highly problematic. Expect results as terse as the first atmospheric braking exercises in the Martian orbiters - a lot of trial-and-error.
Starting with Deep Impact: This prediction is spot on, although it would have helped if I would have quantified with numbers - the Deep impact site was barely discernable, a modest 100 meter diameter crater with no only a few meters of discernable depth - about what an impact of an object on the moon would have created.

It is clear from Mark's blogs entries at planetary.org the approach was a trial-and-error exercise. I hoped, as meticulously as he details the mission, he would have reported the nature of the challenge; but he instead focused on how well the software adjusted to the challenge.

Finally Mercury: Yes, they are having a devil of a time modeling the surface, the Newtonian dynamics require a thin skin, flowing mantle very close to the surface. This should have led to mascons much like what we see on/in the moon, but so far, there is little evidence of mascons. So it is curious.

As for the predictions of under-dense peaks and over-dense low-lands; preliminary data seem to indicate exactly the opposite - so it is back to the drawing board.