In the case of new hammers of the original length, I would imagine that
the "paddle" and the hammer would strike the strings at precisely the
same instant. However, as the hammers wear down and receive at least a
light resurfacing, the mandolin attachment will serve as an obstruction
to the normal hammer travel, producing a different sound. By that time,
at least some of the tiny buckskin punchings glued to each paddle handle
will have disappeared, resulting in a wood-to-wood contact and a
"ricky-tick" effect which I find quite pleasing. Ideally, I feel, the
restorer will do away with the buckskin punchings altogether, and when
the paddles are at rest they should be within 1/8" of making string
contact (but capable of going farther if the strings were not there).
In a slow-motion test the hammer shank should then begin to move the
paddle forward just around let-off when the hammer is also about 1/8"
from the strings; momentum should carry both through to strike the
strings in virtually simultaneous fashion, producing a sound far
superior to the paper-clip type of mandolin rail, that incidentally
being notorious as a cause of uneven hammer wear in commercial use.