Power JFETs

What characteristics would make for a good power JFET? 
I think you mean MOSFETs. I don't believe I have ever seen a power JFET.  What is your application?  Switching PS/  Class D amps?  Linear?
No... I mean power JFETs. UnitedSiC is making them and Mouser is selling them. They're rated up in the 30+ amp range. TO-247 packages. I'm holding excitement at bay until a more schooled mind than mine can look at the data sheets, but just there mere existence of current production power JFETs is hard not to be excited about. 
Oh, heck yes!   I'll check them out.  A power JFET, now that is very interesting.   Some years back I read about a company that had patented a solid state tube.  It didn't have a heater but it did have a grid, plate, cathode, and a vacuum in which it worked.  It was made for RF frequencies but I haven't heard from them in years now.  I suspect they folded.  

I did look at them.   Very interesting indeed but they are not linear enough for a Class AB amp.   They are designed for switching power supplies, so a Class D amp may have some applications.   They are not cheap, so they have to bring something of high value to a product otherwise one can do a lot more with a lot less money.Still, I am glad to see them on the market.  Maybe something in the EV world will use them and the price will come down for a Class D amp on an audiophile's budget.
We've been down this road a little before.

I really don't think it's the device's V to I linearity which affects sound quality so much as distortion, bias and feedback choices.

Still, always fun to see new tech come into the arena. :)
I've been talking over at diyaudio about them. You really couldn't use them in a class AB amp because they don't have complimentary parts. They're all N channel. With thier power rating though they'd make for great ouput followers in a beastly single ended amp of some sort. Think SET with a 50 watts rating. 
A unity gain power buffer!!  Yeah, they could sure work for that!
As for the AB amplifier, remember Vacuum tubes are only N channel equivalents, there is no such thing as a P channel vacuum tubes!  It wouldn't be a complimentary symmetrical design but it would be Class AB.

@erik_squires4:  Actually, the V-I characteristics define the devices distortion and bias point.   Loop feedback may drive distortion down to low levels but the initial distortion is caused by the non linearity in the device itself.  All devices are non linear to some degree, some far worse than others.  
Have you seen the new Korg nutube triodes? They're through-hole mounted low voltage devices, two to a package. I think pairing those up with some power JFET outputs would be real nifty! 
More to discover