A couple of clarifications:
The original schematic is protected by Copyright; you can't legally publish a scan. (Whether anyone cares is a different question.) If you have the original schematic and redraw it faithfully, with the same component arrangement etc., that might still be a copyright violation. But if "reverse engineering" means that you reconstruct the schematic from a populated PCB, and decide on your own way to arrange everything in the drawing, that's fine. Copyright protects the specific rendition of the schematic, not the circuit.
Likewise for the PCB layout: If e.g. you remove all components from an original PCB and scan the layout, that layout is protected by copyright. Redrawing the original layout faithfully is probably still a copyright violation. Figuring out your own layout independently, from the schematic, is fine.
Finally, the circuit (or elements thereof) may be protected by patents. If it is older than 20 years, the patents have expired. Otherwise, check the product labels and the manual -- in the US, applicable patents have to be listed. If a patent does apply and is still valid, you cannot use the patented technical design commercially. (No matter whether you redraw it -- you need to make functional changes to it which get you outside of the patent scope.) You can still publish the schematic or layout, but should point out the patent protection. But you cannot take the next step and turn it into a commercial product.
So here's the situation: I made scans of the top and bottom of the original PCB, and created a BOM of the components. I then pulled the scans into solidworks and sketched lines to indicate the locations of pads, vias, copper pours, and traces that I could see. Based on what I found, I guessed that there were four internal layers - a power plane, a ground plane, and two signal layers. Once I finished sketching out the locations for the copper, I saved it as a DXF and pulled the design, 1:1 scale, into Altium. I then proceeded to trace everything out using the Altium tools. I had to manipulate some of the tracks and vias to make things fit properly, and I replaced a bunch of blind vias with through-vias to improve manufacturability. Once I traced out everything I could see, I had to figure out how everything was connected. This was the most time-consuming process and I had to probe every pad and via on the PCB with my DMM in continuity mode. This took me about two weeks, working a few hours at a time every day. I gave all of the nets pseudo-names (i.e. U1-1 for pin 1 of U1, etc). Throughout this process I was drawing my own schematic that matched the layout. Once I was confident that I had probed all of the points on the PCB and had translated them into schematic form, I went back to the layout and routed my own internal layers which I couldn't see from the outside, based on the connections I determined by probing the PCB. Finally, I went back to the schematic and renamed the nets to more logical descriptions which matched the pin names on the individual components.
The schematic is 100% my own work. For the PCB layout, the top and bottom layers are very similar to the existing board (with a few subtle differences), but the internal layers are completely my own. I have not found any patents for this design - it's all pretty much boiler plate, off-the-shelf parts.
Thanks again for the input!
Matt