For a first effort I am impressed. My understanding is that there is really one full time developer on this so not sure if it is fair to compare it to TP which benefited from two technology acquisitions (explosiaFX and DPIT fluids) and TFD which has been around for well over 10 years and still improving. Comparisons to Houdini are also unfair given how long Houdini has been around and I think their development staff has 4 PhD's working on fluids? Not sure, but what I faintly recall was impressive.
I just want to make sure that people's expectations are appropriately balanced because our eyes have been trained to expect results from far more mature programs so the playing field is by no means even. Plus, this was GPU/CPU from the start where my expectation would have been CPU and then a migration to GPU given the complexities of coding fluids on the GPU (look how long it took Insydium to get there and they have a much larger development team).
Also, and I brought this up in another thread, the "look" of the final image really has more to do with the shaders for both the density and temperature channels (black body shaders) than how the simulation generates that channel data in the VDB files. Density and temperature data is created by the simulation engine in the course of solving the physics to make fluids move like fluids. So, if that data is off then it won't be the "look" that bothers you but rather how the fluids move as their motion is impacted by temperature and density (pressure) within each voxel - all things the simulation engine is resolving with each frame. If it moves right, then chances are you have enough accurate data to shade it any way you want. So, the "look" is driven by the shading. The motion is driven by the physical calculations.
If you don't like the motion, then that is a fair criticism but as for me I am impressed by the first release.
Dave