When my badly stored/stacked/strapped slabs have warped badly, I figured out awhile ago that router flattening the kinds of hardwoods I work with is way too much work because I can only make the thinnest of passes each time. I'm mostly dealing with 1300-2700 Janka hardness woods, whereas your average hardwood runs in the 900-1300 Janka range. I've seen some people convert router sled flattening tables to a leveling setup where their Alaskan mill runs on the frame rails and was thinking of doing that myself. But my existing router sled frame doesn't quite work with the mill and bar widths I have.
I'd never been quite sure why it wasn't a good idea to try to level the bottom side of a slab running an Alaskan mill conventionally on it, other than the slab needing it to be raised enough on supports for the saw to have clearance. And the supports need to be somewhat sacrificial in case you skim through the top of them. I never thought it through properly, and was worried something might collapse if the bar came out the bottom of the slab at some points on the cut rather than cutting off a whole sheet of wood. When I was resawing a thick edge piece into a slab, where the middle of the piece was much thinner than the ends, I went through in the middle before re-engaging as I got toward the other end and it wasn't a problem. (So the back end of the cut collapses - if the bar has exited that part of the cut it makes no difference, there's no binding.)
When I thought it through, I realized what you're cutting off on the bottom doesn't matter if it's whole or little sheets of wood being trimmed off. The cut is bound by your guide on top and your bar at the bottom of the cut. To be clear, I'm using a ladder on the top side to level the warped slab like any first cut. Where I did run into some problems with this skim leveling technique was after I botched a cut by thinking I had set both posts on my mill to 5 3/4" height and discovering midway through the cut I set the inside edge at 4 3/4" and thus did a cockeyed cut for the length of the slab, losing much more wood than I had planned to. When I flipped it over and ran my mill on the flat side to level the angled bottom side, the outer 4-6" of the bar were skimming the surface of the wood on the bottom and I noticed occasional violent jerking of the mill as I went through the cut. I think the tip of the bar was catching now and then on the wood surface and kicking back hard. It made for a much uglier cut than usual with a lot of small gouge marks. So I've concluded you really want the outer end of the bar buried in wood and not skimming when doing any leveling. Other than that it seems to work fine.
The table below with 4x4 blocks on top is my resaw table and that's a partial sheet of red oak I trimmed off. Usually I move the blocks around on the table and screw them in at the bottom, to chock short slabs I'm resawing (there are lips on top of the blocks). When doing larger heavier slabs which don't need chocking as much like the ones I was skim leveling, I just place them on top of the blocks. I did slide the cockeyed slab off the rear block while leveling the second side and I didn't bother putting it back in place and finished the cut going at an upward angle and it made no difference to the cut. More proof that the mill being guided properly on the top side is the only thing that matters.
I'd never been quite sure why it wasn't a good idea to try to level the bottom side of a slab running an Alaskan mill conventionally on it, other than the slab needing it to be raised enough on supports for the saw to have clearance. And the supports need to be somewhat sacrificial in case you skim through the top of them. I never thought it through properly, and was worried something might collapse if the bar came out the bottom of the slab at some points on the cut rather than cutting off a whole sheet of wood. When I was resawing a thick edge piece into a slab, where the middle of the piece was much thinner than the ends, I went through in the middle before re-engaging as I got toward the other end and it wasn't a problem. (So the back end of the cut collapses - if the bar has exited that part of the cut it makes no difference, there's no binding.)
When I thought it through, I realized what you're cutting off on the bottom doesn't matter if it's whole or little sheets of wood being trimmed off. The cut is bound by your guide on top and your bar at the bottom of the cut. To be clear, I'm using a ladder on the top side to level the warped slab like any first cut. Where I did run into some problems with this skim leveling technique was after I botched a cut by thinking I had set both posts on my mill to 5 3/4" height and discovering midway through the cut I set the inside edge at 4 3/4" and thus did a cockeyed cut for the length of the slab, losing much more wood than I had planned to. When I flipped it over and ran my mill on the flat side to level the angled bottom side, the outer 4-6" of the bar were skimming the surface of the wood on the bottom and I noticed occasional violent jerking of the mill as I went through the cut. I think the tip of the bar was catching now and then on the wood surface and kicking back hard. It made for a much uglier cut than usual with a lot of small gouge marks. So I've concluded you really want the outer end of the bar buried in wood and not skimming when doing any leveling. Other than that it seems to work fine.
The table below with 4x4 blocks on top is my resaw table and that's a partial sheet of red oak I trimmed off. Usually I move the blocks around on the table and screw them in at the bottom, to chock short slabs I'm resawing (there are lips on top of the blocks). When doing larger heavier slabs which don't need chocking as much like the ones I was skim leveling, I just place them on top of the blocks. I did slide the cockeyed slab off the rear block while leveling the second side and I didn't bother putting it back in place and finished the cut going at an upward angle and it made no difference to the cut. More proof that the mill being guided properly on the top side is the only thing that matters.