A new MSX (3?) for 2022 ?

Thank you for sharing this information. If these parts are indeed still available then it would solve one critical variable in the compatibility equation.

As a fan I definitely look forward to the MSX3. However it It is still not clear to me what MSX3 is supposed to achieve, or what it is supposed to compete with? How will it take PC forward? Is it simply a personal project, or a fan service? Or much more than that?

ARM cpu is the realistic way to go if we want a modern continuation of the MSX standard.
Using 20-40 years old z80 technology to create a new MSX standard is Only for the few nostalgic People who would buy that. For new users a z80 compatible cpu means low speed and high cost and barely upgrade options.
So z80 would be DOA in terms of commercial success.

Nishi stated several times that focusing on retro is not the way he wants to proceed.

"High cost" seems clearly wrong to me, since the Z80 and eZ0 run around $8-$15 qty 1, which is cheaper than any FPGA you'd program to do the same thing and not much more than the low-end ARM CPUs. Nor is the speed uncompetitive, at least for the 30-50 MHz eZ80s, unless you're trying to run a modern OS. What are the upgrade options you're thinking of that couldn't be put on a modern Z80 or, especially, eZ80 system?

I'm not really clear how an ARM system is a continuation of the MSX standard, though. If you go this direction, and use emulation to be compatible with MSX1/2/2+, what's the difference between this new "MSX standard" machine and an $80 Raspberry Pi running Linux and emulation software?

I would say that anything trying to continue the MSX standard is going to be a commercial failure compared to something like a Raspberry Pi, and appeal only to those who are nostalgic for MSX. This seems very much a niche market.

As far as success on its own terms, I'd expect price would be the biggest factor. If you can sell a machine for under $100 you'll probably pick up curiousity seekers and the like, but a $700 unit is going to appeal only to the very few who really want an "MSX" thing that runs a small fraction of the software you can run on a PC for half the price (which, keep in mind, can also run MSX emulators).

Konamiman (Nextor OS coder) presented, in Japan, msxVR project to mr. Nishi ;
then opened his eyes and mind ; Nishi liked that project and wanted to make something similar.

mr.Nishi: please make your new computer: turboR compatible (msxPlayer was not) (neither 1chipMSX) and VR compatible. You must talk with VRteam

SCC and v9990 are MSX standards nowadays.

I like FPGA projects: MiSTer, ZX spectrum Next and Mega65.
I like ARM Mini Classic machines.
And besides, I like very much RaspPi side like msxVR.
I'm proud owner of a MSX tR and MSXVR.

I like MSX scene.

Hello, i'm lurking the MSX scene from some time (because i had a MSX1 computer as a kid and having fun seeing what people do with it nowadays) and decided to post my opinion.
Difficult to understand what people can expect when there's very little info. The most accurate one can be the hardware used (not sure if it's final) and what you can possibly do with it and with that try to imagine what can be realistically done. And from that what can realistically be popular.
From the photos posted we know there's an ARM processor with FPGA functionality. We know it will run virtualized hardware for backward compatibility like other emulation solutions (MiSTer FPGA). But the question is how the MSX3 "mode" will work.
If they try mimic the Raspberry i don't think they can create an entire OS with desktop functionality by themselves. I'm not expert about how FPGA systems work on a software level but maybe in the end this MSX3 could be a linux system with virtualized hardware trying to mimic MSX programming paradigms on the side (ex. API for video, audio, etc). I'm not seeing enough manpower to create (and maintain) a new OS and desktop if they want to provide modern common functionality like browsing and so on.
If the intention instead is a middle ground between devices like Arduino and Raspberry it's more interesting in my opinion: a small device with simpler software architecture but with powerful video, audio, input, output functionality compared to Arduino. Instead of having to "fight" with the linux enviroment to do something simple in Raspberry you could have a simpler way to approach the various components of the system. For example i created a surveillance system with Raspberry but having to strip linux distros and understanding how things work was not easy, while with Arduino the hardware wasn't powerful enough to do what i wanted (or with too much expensive boards to add together).
An ideal MSX standard would provide simple components to access with clear documentation. That's an important thing because it can be really what can make a difference in potential customers. So many additional boards or clones in IoT have barebone documentation and they can be difficult to use. If a new MSX system can provide a more reliable programming layer to people it would be nice. That was useful then (in the 80s) and it's good now.
About CPU, video and audio capabilities i don't really know what the ideal approach would be. It's all about how simple to program or what you want to run on the hardware. Would you provide modern GPU paradigms or stay on classic VDPs of 8/16 bit computers (programming wise of course)? Maybe with FPGA you can have multiple options: being able to port software from other systems (ex. OpenGL or Vulkan APIs) or use a virtual rendering system with simpler API (ex. focus on 2D or simple 3D) with maybe easy builtin functionality (ex. collision systems, debugging).
Adding, extending the hardware and software without locking to a precise manufactures or goal is an important factor future-wise. I see why the board itself is a custom one but ideally i'd like to see focus on developing MSX more on a software/engineering/standardization level and let people create their systems with multiple options.