It's a difficult one to work out volume of expected sales vs cost of R&D and assembly / retail price etc.
There are simpler (well, I use the word simple in the loosest fashion since in practice it might not be so simple lol), but suggestions from me (and I have mentioned a few of these before) are:-
1) Sega CD Drive Solution
2) FPGA VIC-II replacement - outputting RGB.
3) PC Engine HDMI / DVI adapter - something inline that can use the VD bus etc to generate a better than RGB output.
4) C64 / Spectrum / smart joystick adapater. eg. on the C64, could connect to the keyboard connector internally, connects to the joystick ports too, has 2 neo standard 6 button stick ports, and allows the user to configure keyboard presses to additional buttons. could be simple to design, could be hard depending on whether its generic or different models for different systems. Could be popular though - I know its a pain the ass trying to hit space or some other keys to try and play games using a stick on the C64, Amiga, ST, Spectrum etc. I could do something like this easily with a PIC - might have a go for the C64 one day :P
5) Atomis Wave Multicart
6) CPS1 & CPS2 100% emulation device (could perhaps be extended via firmware updates to support other systems), OR some kind of ROM board that will allow the CPS1 or CPS2 to play all games of that platform.
7) An affordable scan doubler that outputs DVI or HDMI. I know there's the framemeister and OSSC etc, but they still feel like complex devices. We need something that can accept SCART, RGB, S-Video for all the different consoles that 'just works' with easy to use interface. Maybe the OSSC is that device, just looks a bit 'kit ish' to me, but maybe that's cost effective and I imagine accepting all sorts of video input variations a complex hurdle to overcome.
8) Amiga CD32 / CD emulator device to give an A1200 CD drive support (ISO etc in first instance), maybe later with Akiko support or whathever hardware the CD32 provided over the 1200 - but that's a complex bit of kit imo. Sales could be high though.
9) Generic optical drive (CD / DVD) emulator - could be utilized in all sorts of system, including in the first instance the Commodore A1200 / A600 - easy to just plug into the IDE chain and use an SD (like you've done with the PCE). Could also be used with any other IDE device that has a CD / DVD drive. Could be cheap to design, could be popular due to the generic-ness of the device.
10) Sega Saturn / Dreamcast / 3DO drive emulators. Some already available yes, but hard to buy as always low stock. 3DO would attract few sales, Saturn and Dreamcast would sell a lot I imagine if they were always stocked atm.
11) Neo CD Drive Solution - not sure very high demand there, but could be easy to implement, might be very similar to point 9 in that once an ATAPI drive exists, it might not take much to convert that to work with the Neo CD - maybe a change from ATAPI to whatever SNK went with on the Neo CD, but point 9 could lead to point 11 or vice versa.
12) There are also some possible easy wins that could help bring small amounts of regular cash into the company - ie. products that are open source, that could be just tweaked, manufactured and sold too. eg. RGB amp PCB, stereo sound amp PCB, PSU replacements - maybe even a generic PSU that has multiple cable options. People are always looking for new PSUs for all those old computers and consoles. Terrible Fire Amiga accelerators (anyone actually building and selling these - I am looking to buy one, maybe £150 without an 030 on board? I bet cost of components is like £60 or something. The chap over at Terrible Fire (You Tube) has already said people can build and sell them, just dont over charge etc. But no one to my knowledge is harnessing that as a possible way to make income and to help people wanting to just buy one ready assembled. Despite the fact the Vampire is faster and better, people also just want an 030 or 040 on their Amiga at reasonable price (even if the 68K is emulated), and Vampire is expensive, and the 50Mhz cards from Individual Computers are crazily expensive.
13) I think a MAME box might sell well if done properly, but there's the Pi, XBOX etc that competes with the general idea there - that's the problem with most FPGA 'system replacements' imho.
14) Re-seller / licensing for all the great projects that have low volume due to hobbyist type projects etc. examples might be Neo Geo ROM adapters (for boards without a DIP ROM), AES Save Carts (both from Turfmasta I think) - these might be always stocked these days - no idea, just remember that whenever I want something I am unlucky and missed a batch etc). Amiga ROM switchers, C64 ROM switchers, bwacks 5v 'saver', Dual SID boards, Master System FM board or something to add FM to Game Gear / Megadrive / Master System (Again often out of stock with people making these). NES / Famicom stereo audio kit, Spectrum +2 / +3 Stereo Audio Kit. The list goes on, and a lot of these are cheap to make, probably wouldnt sell for masses but volume of cheap low risk products that combine with all other product offerings could help keep things ticking over. Some products could compliment your range too - eg. MVS to AES adapter (Furtek or own design?), PSU for AES (9v and 5v models), AES AV leads, MVS Consolization Kit (eg. 5v PSU, controller ports, AV DIN and AV SCART cable, audio amp PCB, optional licensed unbios). There's a lot of scope there, but granted the last thing you want to end up doing is spending loads of time on too many low volume products.