Atlas O track, I'm told in another forum, uses code 215 rail. My quick check with a ruler (not calipers or micrometer) said around 1/4" or code 250, so I'm willing to believe code 215 (double the rail height of HO code 100 track).
The prototype normally built turnouts and crossings with heavier rail than the lighter track used, usually the same as, but sometimes heavier than the heavier use track. So it would be appropriate to transition the On30 (HO gauge) track to the code 215 before getting to the crossing. That would take a couple of inches to transition that 1/10+" difference in rail height.
Thinking about the problem some more - wiring the bugger would take a little planning. If you allow the O27 center rail to be discontinuous and insulated from the rest of the rail in the crossing, you would only have to have gaps in the 2 outside O27 rails between the 2 frogs. All the rest could be continuous metal rail and frogs, as far as the On30 is concerned. However, there would be short circuits when O27 center rail rollers were touching both the crossing On30 rail and the live center rail at the same time. So the On30 rails would have to have a short isolated section where the O27 center rail would be. Then insulate the whole crossing from the rest of the track and use a DPDT to select whether the crossing would be set for O27 or On30. When set for On30, all the On30 track segments would be live, and the O27 center rail would be dead. When set for O27, the center rail segments and On30 center segments would be live, and the rest of the rails would be connected to the O27 outer rail.
After sorting through the wiring issues, I realized that the center O27 rail can be soldered to the isolated center section of the On30 rail with just one gap at the center of the crossing, since it will be switched as a unit. This would give an H configuration of metal rail down the center of the O27 path. As long as the top surface is kept smooth, should work like a champ.
My stuff is in storage while I'm doing a drawn out move so I don't know how wide an O27 flangeway has to be. Except for that worry, it would be a fairly simple exercise in handlaid track construction as long as you are willing to use soldered frogs and add all the gaps and feeders needed for the above wiring scheme.
Still, I would think an over/under bridge crossing would be more reliable and fun if the room can be found for the grade - rail to rail clearance would have to be about 5.5" for the On30 to go over the O27; maybe an inch less for the other way around?
An interesting planning exercise no matter which path is chosen. I had done a feasibility check for the minimum size for an O27/On30 layout that featured a narrow gauge to standard gauge transfer facility as its focal point. It fit into a 5ft x 8ft space with an O36 oval with passing siding and 3 spurs, one of which served the transfer platform. The On30 used 18" radius on a small point to point (half oval) from the transfer platform to a town at the top of a grade.
I think the mixture of O27 and On30 would be a lot of fun, as long as you are not taking prototype modeling too seriously. Most O27 and On30 users don't because of the obvious discrepancies in their track, so it shouldn't be an issue. Now that I've done so much planning of the crossing, I'm going to have to build one for kicks when I finally get relocated!
enjoy