Glad to be part of that annoying sense of guilt at the back of your mind :)
For me it's less about modeling the tube and more about checking the rest of the circuit around the tube if that makes sense. Does the tone control do what I want, and so on.
Personally I'm looking to verify relatively simple parameters of a tube in a circuit: gain, whether it's operating in the linear region given a particular bias, that sort of thing. If the model was simply a representation of the tube's characteristic curves that'd be good enough for my purposes.
So being SPICE-like would work ok. After all that's what I'm doing now!
My hesitance has mostly been my inability to verify the simulations. I prefer to have physical elements available so I can double-check the app. But I think I will have to resign myself to verifying against SPICE or something.