Hi!, about the track/car data storage, you are right, if it stores every lap of every session in the long run the prediction would be off. Also the fact that during one session you could use one setup, but in the next maybe you use another setup, this can makes a difference in consumption between sessions.
But an average during a whole session i think it's better than just the average consumption from a few laps in practice, because it will be more representative of the true consumption during the whole race, as it includes the moments where you are going to probably push harder than in practice.
I was thinking this: session Nº1 starts, the first with this track/car combo, so there's no previous data stored and the app calculates it for the first time. When closing this session, CBV stores a "last session" consumption and lap data. Then, you run session Nº2, and CBV allows you to see the numbers from the last session, but begins to calculate new numbers for this session. Then, you close session Nº2 and the data from session Nº1 is erased and the data from session Nº2 stored instead. You start session Nº3 and CBV shows the calculations from session Nº2 and independently calculates new numbers for session Nº3, and so on. This way, the moment you open a new session you can have preliminar numbers that won't represent a big average (with poor precision) from hundreds of laps, instead they will represent just the calculation from previous session, and you can also do a new calculation for the current session and work on it if you have the time to do it. Also in this way you prevent the files that store the combo data getting bigger and bigger over time in an unnecesary way.
This is the way i would think it's a good way to implement this feature.
