This is Lionel:

Actually your question is a lot more complex than at first considered.  I
would have to ask first, "Is this CHEMICAL or PLASMA etchback?"  There is a
lot of difference between the two.  I would refer you to a round robin
performed by ITRI a few years back; the #1 result was obtained using
plasma.

I have seen a lot of problems with chemical etchback, particularly with
ultra-small holes.  The problem has been either residual chemistry
remaining, or incomplete etchback, due to mass transfer problems.

However, the specific energy choice for plasma is also a problem; mega
versus giga hertz systems produce radically different results.

As to answers, I find that flex or polyimide manufacturers tend to be much
more secretive than standard rigid board people.  Not much in the
literature, I am afraid.