I concur. My thoughts after checking out the data sheet:
General specs
It's a directly heated "tube" and at 0.7V filament voltage requires DC heaters. The given transconductance figure must be microhmos because the plate resistance is listed as 330k (though Mu = Gm x Rp and it doesn't quite work out).
Charts
There are two I/V plate charts: one at low voltage with positive grid bias and another with a larger voltage scale. With the 1.7mW plate dissipation, this is definitely designed for a positive grid. The curves below the dissipation max at higher voltage (where the grid could be biased negative) are kind of ugly. The curves look most interesting on the large scale chart in the 20-60V range, but the 0V grid line is smack in the middle of it.
Apps
This all makes sense given that it is direct heated and will likely be paired with other solid state parts and low voltage regulators. The app notes have a clever bias scheme for a 3.3V supply (note this will only work with positive grid bias). The 'Basic Circuit' shows source followers on both the input and output to deal with the impedance issues of the positive grid and large plate resistance.
Frequency Response and THD
Frequency response looks good at the voltage supply levels provided (5, 12, 30) with a -10dBu input (about 0.25Vrms). THD+N is not so good at the 5V supply level, but it is not bad at the -10dBu input and below with 12V or 30V supplies. Looks like it's meant to be driven with fairly modest input signals (ie portable devices). There's no indication of the harmonic make-up of the THD on the datasheet.
Conclusion
This is all about knowing the limitations of the device and what the trade-offs gain you. This "tube" is an interesting hybrid in terms of specs and it definitely was designed to be partnered with solid state. I associate Korg more with MI than CE and so I was expecting something destined for guitar preamps and pedals (and this might work fine for that given there are two triodes in an "envelope"). However, it also appears to be an interesting candidate for a gain stage in portable headphone amplifiers, given the low supply voltage requirement and the low output voltage limitations of many phones and DAPs.
I don't get the desktop format in this particular amp other than that it makes for a fairly unique and interesting device. If portability isn't a concern, there are better "tubes" to use. On the other hand, the Nu Tube device itself may allow the manufacturer to keep some production costs down in comparison to traditional tube amps (all on a PCB, no high voltage or output transformers, consistency from device to device). Whether it's better than a JFET in the same application would be interesting to hear.
In all cases, this guy needs assistance from ancillary solid state devices. Ain't nothing wrong with that, but given this requirement it's almost more of a Nu Transistor than a Nu Tube.
