I’ve been fiddling with different methods of rust removal since being bitten by
the rust bug.
Chemical methods have advantages. Instead of rubbing and scrubbing to get the
effects of oxygen off of my iron and steel the use of chemicals like citric,
hydrochloric, and the case of vinegar acetic acid, acts on the surface rust
almost effortlessly. But like all really easy options there is a catch. The acid
will dissolve the rust fairly quickly. When the rust surface is removed the acid
starts working on the iron and steel.
I was cleaning a saw set and left it in the vinegar overnight. The frame of the
set was nicely cleaned along with the adjustment screw, but when I went to put
the thumb nut back on the screw it rattled.
I have a theory that I believe explains both the screw being smaller and the
more useful effect the acid has on the sharpening of files.
Since I believe the acid works on all surfaces equally the acid removes an iron
atom from every surface it touches. The screw thread, and file teeth, are
triangular is cross section. Since the acid is working on the all surfaces the
two sides each lose an atom and the point loses one also. That causes the
thread, and tooth, to get thinner twice as fast as it gets shorter. Good for
file sharpening not so good for cleaning screws. The file getting thinner
doesn’t affect its function. The screw, on the other hand, has a problem when
the diameter is reduced. Not only does the diameter get smaller, the thread is
getting narrower. Next thing you know the thread is too small and ruined.



What does this long explanation have to do with pits? 
Like a tooth, or thread, pits have two sides so they get wider twice as fast as
they get shallower. Mechanical methods only work on the surface with the result
that the pit can only get shallower.
I think the same thing happens with any surface imperfection. A spring with a
crack will get cleaner and have a bigger crack when the process is done.
Add the effect of packing a bunch of hydrogen atoms into the iron carbon matrix
and the spring ends up inflexible. I think that any weakness is multiplied by
the chemical method.
When I dealt with armor plate it was imperative that the metallurgy remain
unchanged by the process. During the painting process I could use a chemical
pretreat on 12560 with no concerns. 12560 is tough. MARS300 is hard. It couldn't
be pretreated because it would crack 3 days later. Hydrogen embrittlement was
the cause.