The advertisement heading is a bit misleading- the LeRoi is in a power unit.

I've always found this for LeRoi as a marketing manufacturer, but they were NOT shy about producing products for the other companies, nor were they married to internally-sourcing everything.  I would not be least bit surprised if there was a significant manufacturing contract between AC and LeRoi, particularly in these 'small' engines (LeRoi's 'big' engines were significantly larger)... and keep in mind... Allis-Chalmers was making engine-driven generating systems where THEIR specialty was the GENERATOR end...

The flat combustion chamber, in conjunction with a dished piston, was actually an 'inverted hemi'... and to AC's design merit, a much EASIER design to produce accurately, than trying to shape a cylinder head with a 'combustion chamber'.    Engine designers had a heckuva time getting castings to provide a consistent chamber shape cylinder-to-cylinder... but it's easy to get a consistent piston dish, and VERY EASY to get a flat cylinder head with two valves and a spark plug in it.  I don't recall seeing an Allis patent reference on that concept, but my gut says they probably did, and they probably guarded the heck out of it, as it yielded one of the most balanced (cylinder-to-cylinder power wise) line of engines out there.

I'm certain that LeRoi's LARGER engines was doing (with their chamber design) what Waukesha's larger engines were doing too-  running more chamber volume, changing piston height and dish, to get a final CR that was appropriate for the fuel type.  Frequently, an industrial engine manufacturer who did this, was expecting they'd run natural gas or propane, under higher compression ratio... or swap the head with a much smaller chamber, or a precombustion  (like a Lanova) design.   For gasoline or gaseous, a Ricardo chamber design would be the choice, and I think this may have been what LeRoi might have been aiming for...  but this is simply historical speculation on my part.