Nat,
I figured after Dave Hillman posted his response that the question would be
coming, so I started to look over the situation.

1.  Since you have all those PTHs, you really have a mixed technology board,
not an SMT board.  Is it truly a double sided board or a multi-layer?  This
can make a difference if we are talking about entrapped flux.

2.  Personally, I like LPIs.  When I was with Naval Avionics, I cursed dry
films and the demon-on-earth that invented them.  I have heard from some that
there are LPIs which can tent vias, but I don't have any of the formulations
which supposedly can do so.  Perhaps some of the other TechNet individuals
could help in that respect.  Use a high quality mask though.  One that will
withstand at least 4-5 passes to molten solder.  Anything less and you risk
the situation of a cheap mask being highly absorbing of HASL fluxes (can you
say Metal Migration?!).

3.  I think you are wise to be concerned about the HASL residues.  I have
lost track of the number of no-clean assemblers we have helped that have been
bitten by HASL residues on the bare boards.  Find out from your fabricator
what HASL flux or fusing fluid they are using.  Is it mostly
chloride-activated, bromide-activated, a mixture of the two?  What is the
cleaning operation?  In-line, batch, tap water, DI water.....?  That kind of
information can better help me determine how likely a fabrication process is
to leave detrimental residues.  

4.  In most cases, HASL flux residues are more in the laminate and mask, the
areas between traces, than in the PTHs themselves, although that is a
possibility.  Your assumption that most of the HASL flux/fluids vaporize
during HASL is only partly correct.  Part is vaporized, part is absorbed.
 Usually, the higher the solder mask quality, the lower the absorption.

5.  You state that you are using a leave on flux, which I would term a low
residue flux or a no-clean flux.  Is the flux halide free?  I would recommend
it be so.  Some flux vendors have marketed no-clean fluxes which contain
halide elements.  Not a no-clean flux in my opinion.

6.  Ion chromatography (see IPC-TM-650, method 2.3.28) may be able to do what
you are looking to do.  Our internal method is much more detailed than the
TM-650 method, but still covers most of the board surface, not the individual
PTHs themselves.  IC is the only method readily available that might be
sensitive enough to detect the low levels of corrosive elements.  Send me a
couple of the boards and I'll see if I can adapt our method to PTH only.

7.  Can you share what the assembly will be and what end-use environment the
assembly will face?  Jungle environments drive corrosion and metal migration
far more than office environments.

Hope this was helpful.  
Doug Pauls
Contamination Studies Laboratories (CSL)