Add rationale information to Projectile for doffset constraint.

[smiths]

As implied by #2264, the rationale for the input constraint on doffset is not explained in Projectile. In IM:messageIM we are told there is an input constraint that doffset > -pland, but this is not justified.

Should the justification be given as part of IM:messageIM (as an additional derivation), or should it be part of the Properties of a Correct Solution section? We should give this some thought and then change the Projectile example accordingly.

[JacquesCarette]

It depends on whether we view that constraint as part of the domain description or part of the description of good solutions.

It looks to me like this isn't a constraint on the problem itself, but its solution. But the description above assumes enough knowledge of the details of Projectile that I could be wrong.