I'm sure it's just the DT midi conversion routine; there are a few where the music doesn't start until a few hundred bars in. The mirror site version is perfectly ok. I'd render the abc slightly differently to get the beams and the triplet:

X:1
T:Fanny Blair
B:Gale Huntington, Songs the Whalemen Sang. Barre, 1964: pp. 229-31.
S:Text: Log/journal of the ship Java, out of New Bedford, 1839. Tune: source unspecified.
N:Huntington prints the two final notes in bar 7 as eighths (quavers).
Q:1/4=100
M:4/4
L:1/8
K:D
(D/2E/2) |FF FE/2E/2 F2 DE |FA Bc d3 c |
w:Come_ all ye young men and mai-dens, where e-ver you may be, Be-
(cB) Bd (cA) F^G |((3A^GA) BB B3 (A/2A/2) |
w:ware_ of false swear-*ing and sad__ per-ju-ry; For it
A^G AB (BA) F A/2A/2 |(Bc) dc B3 F/2=G/2 |
w:is by a false wo-*man I am woun-*ded so soon, And you
AA BA (FE) F D/2D/2 |(EF) DD D3 |]
w:see how I am cut_ down in the height_ of my bloom.

The two final notes in bar 7 are changed to sixteenth-notes here, which fits the time signature; in Huntington they are eighth-notes, which is either a mistake or the bar is in unindicated 9/8 time. I don't know which is right, as I don't know what the source for the tune was.

I notice that the Wyman-Brockway set was collected in Letcher County, Kentucky, which is a little unfortunate in the circumstances; or apposite, depending on how you look at it.