There seems to be very little, but then again, a lot of the little scraps of practice in things like Drinker's letter or the Martha diary would be incomprehensible to us without a more full picture of iron operations elsewhere. There's an account of practice at Nassawango Furnace on the Delmarva, which operated in the 1830s (
https://books.google.com/books?id=6GdDAQAAIAAJ&pg=PA314) and used ore-raising methods that sounds similar to those employed in the Pines. If I interpret it correctly, they roasted the "hard" or massive ore to break it up, but perhaps not the softer loam ore. Accounts of practice at the end of the period (e.g., Overman's treatise) suggest that bog ore should always be roasted, but who knows whether this was always carried out.
The Martha diary for March 24, 1811 has the entry: "All gone well. Ore heap afire. Ore teams carting sand." The second sentence suggests a roasting operation to me; the diarist doesn't make it sound like a catastrophe (the sand is for molding).
My impression is that Gordon's caveat about needing ore of a minimum 55% iron content for successful bloomery smelting probably applies to bloomeries in general, which seem to have consumed higher-grade ore and of necessity produced fairly iron-rich slags. Given the general fussiness of the blooming process, I suspect roasting would have been necessary in general for bog ore fed into a bloomery forge in order to produce anything like consistent, quality iron, but this is speculation on my part. The simplest form of roasting would have been simpler than charcoal burning: just make a gridiron of logs, heap ore on it, ignite and allow to burn (no turf necessary)!
I know in the Piedmont and Ridge & Valley, old charcoal flats can be detected on LIDAR from their alteration of contour and field-checked by the detection of a charcoal layer in the soil; has this been done in the Pines as well? Presumably a roasting site would look similar but leave a mixture of charcoal and powdered ore somewhere in the soil.
Note that the high iron content of bloomery slag (presumably the reason Drinker condemned the method in favor of furnace + fining) meant that it could potentially be reclaimed in a furnace charge. The stamping mill at Martha (where Moses Gaskill lost his finger after a month of working) could have processed the furnace's own cinder, and broken up massive ore from the bogs in lieu of, or in addition to, roasting; but one wonders if bloomery slag from elsewhere in the Pines was carted in as well.