I think you can figure this one out.

Any time there is a bond-breaking step, there will be an energy barrier. So, yes.

Yes, secondary alkyl halides should be the most common case where you would observe carbocation rearrangements.

Hydride shifts, generally, will perform 1,2 shifts much faster than alkyl shifts. First, think about the driving force for a 1,2 shift. The driving force is formation of a less substituted carbocation. If you have a secondary carbocation adjacent to a tertiary carbon, and the tertiary C-H migrates, you obtain a (more stable) tertiary carbocation. However if an alkyl group migrates, you obtain an (equally stable) secondary carbocation.

Exceptions can occur. In the lab, there are cases where 1,2- alkyl shifts will happen preferentially, especially in rigid cyclic systems. In rigid, cyclic systems containing a free carbocation the groups likeliest to migrate are those which have their bonds aligned with the empty p-orbital; axial groups, in other words. See the references at the end of the post for some examples.