They should never have killed proximity faults.
At least with what I use now my fine work comes in as 2D and 3D when I run field to finish, so I do not have to do the dance you outline above.
They didn't kill proximity faults, they are still in there. However one might still
forget to add any desired weeding or supplemental factors to them when they are added to the TIN. And then, there you are doing the Breakline shuffle all over again.
Then again I recommend strongly against their use. For this reason, if there should be a break or a feature that should have a breakline to define it, then the surveyor should collect data to properly define that feature.
And by definition a 'proximity' fault does not.