Super Mario 3D World + Bowser's Fury

The first thing is: 8 data points is a lousy set that's dramatically shifted by outliers, so any predictions need to be really broad.

A simple average is a lousy metric to use because it gives no weighting to games that might have sold more or games that might be more relevant to the current case.  At the very least you want to compare against a weighted average, which is total over total: removing BotW (correctly) that's 211% extra, which still suggests that something's wrong because it's bloody high compared to the nearest comparisons.

Next step: checking the data.  BotW is obviously an outlier because it released at the same time on both, and Mas correctly excluded it, although that 305% shouldn't even be near the final results.  However, you also have Mario Kart 8 which was released in the first fervour of Switch frenzy, and (anecdotally maybe, but industry knowledge is important) picked up several purchases among those who now wouldn't touch a minimum-effort Wii U port, like me.  Exclude.  Bayo 2 is a very different sort of game to SM3DW, much nicher; and arguably so is Hyrule Warriors.  I'll at least be excluding Bayo 2.  That brings our simple average to 118% and our weighted to 132%.

This is going to end up being very similar isn't it.

Are there any games we want to give more weight to?  Captain Toad and New Super Mario Bros. U feel like the closest to SM3DW, so give them double weight in our averages.  Simple average becomes 120%, and weighted becomes 133%.  You could arguably reduce the weight of Hyrule Warriors too, but that has minimal effect anyway.

So what have we got?  133%, which isn't too far off, but still a relatively large spread among our 5 games.  Looking at them, about the only thing I can say with confidence is that we'll see between 100% and 150% extra sales.  6m to 9m, then, at £40 a pop.  138%, even if it lands near my calcs, is far too accurate a number for what data we have.