I think the greatest factor to the lack of Pyra sales...
*ahem* You mean "pre-orders," yes? Thus far, 0 Pyras have been sold, considering a final revision doesn't exist. Ergo, no Pyras are available to purchase. Besides, a pre-order can be cancelled at any time, for any reason, without a single Pyra leaving Dragonbox.
Okay. Let's say, for comparison's sake, we count Pyra pre-orders as sales (which they aren't). What does the GPD WIN have over the Pyra besides:
A x86 processor [which is] much more useful than an ARM based one overall.
Well, the WIN had a hugely-successful IndieGoGo fundraiser, which earned it a budget that's at least double the current funds raised by the Pyra pre-orders. Funds raised by a very broad audience made up of far greater numbers than the Pandora/Pyra community.
Successful campaigns on such sites are also very potent for marketing and promotion. How much promotion did the Pyra get to this point, exactly? Besides the very small amount of marketing and promotion this community alone could provide, there's no way the Pyra has as much wind in its sails as the GPD WIN does right now.
If this pre-order had happened several months earlier, then we might be in a different situation.
Well, they tried to make a robot EvilDragon that runs off of nothing but coffee in order to get the Pyra into prototyping phase much quicker. I wonder what ever came of that project. Hmm...maybe they scrapped it to work on getting the manufacturing process straightened out. Who can tell?
Or...maybe...the robot EvilDragon developed a mind of its own, killed the original and took over! *gasp* Kill the Synth! Down with the Institute!!!
If there's a cpu board released with an AMD or Intel x86 SoC then I think we'll see quite a large increase in interest.
If only the Pyra had a Neo-Geo coming out of the 2nd SD card slot...or a Thunderbolt port to hook the Pyra up with an nVidia Quadro GPU...or a manufacturer of power-efficient x86 SoCs that is willing to sell their silicon at a very low MOQ.
One can dream...
A secondary factor is the quality and production problems the Pandora experienced.
Both of which are mutually-exclusive to the Pyra, which is developed in a different way and headed by a different person.
And here's where the Pyra's lack of marketing force can come into play: building the story. Market the Pyra as the successor to the Pandora, but better in how it's made.
"The Pandora was burned to the ground by a madman who jumped off of roofs to promote his gadgets. From its ashes, the Pyra rose. Led by a man who breathes technology and bleeds coffee, never will the Pyra suffer the same pitfalls in its development that plagued the Pandora."
What I'd like in the future is a dual boot Pyra Running both Mint or Arch and Windows on a 14nm AMD SoC. I don't really see it happening, but a man can dream can't he?
[RANT] Mint is one of the most bass-ackwards distros I've ever seen. "Hey, let's fork GNOME3, and never update it! We'll call it 'stable!'" Ever heard of
Solus? They got it right. Focus on what matters the most: the UI. Gnome's got the underlying tech squared-away. No need to fork every component under-the-sun and give it delicious-sounding nicknames.
[/RANT]
Well, if AMD were willing to sell at a low MOQ, then it's a possibility.