I'd say that has more to do with Nintendo not really deciding on what Switch's post launch year would be yet internally. Anything EPD was working on prior to the Switch's launch was strictly for the console's first year, and new titles green-lit afterward were only done once the console became a success. As Nintendo likes to take their time with games, next year will be when more big EPD projects come out, same for 2020 onward. So in a way, you're correct that Nintendo burned through a lot of their resources, but that's because they were entirely focused on making the Switch's first year the best it could possibly be. That included getting BotW and Odyssey out complete and on time. 

Oh yeah, they definitely fired a lot of their big guns last year, leaving this year somewhat light on big releases as a lot of their key studios "reload" to fire again in 2019 and 2020. Even so, they have to be careful how they allocate their finite resources as compared to MS or Sony they have to do more of the heavy lifting for themselves in terms of driving hardware sales. 

They should of made the current title a more service like game. Even though it isn't like Nintendo to do that, it could work with the way the game just is. More profitable and cheaper than making a completely new game like we think is happening.

1. So you're using VGC as the source for SFV's number but Nintendo as the source for ARMS' number. Why?

2. That's the thing. I'm not trying to debunk anything. I'm asking for a source proving your claims.

3. Usually, the burden of proof lies with who declares (you: "ARMS has outsold Tekken 7 and SFV retail sales").

4. I thought you knew that Capcom doesn't mix sales numbers of re-releases together with sales numbers of original releases. SFV and SFV Arcade Edition as well as SFIV, Super SFIV, Super SFIV Arcade Edition, and Ultra SFIV are all separate. As are every other original release and re-release.

5. Yoshinori Ono in December 2017: "Accumulated lifetime sales is getting close to 2.5 mill."

Even though you can't seem to back it up with a source, I'll give you that ARMS could very well have outsold SFV at retail. This literally cuts out the PC version though as well as a chunk of the PS4 version and one may wonder why it's necessary to set up these terms in order to make ARMS look succesful. Considering that SFV is a game that many have deemed a huge flop, I fail to see how any of this proves that "ARMS exceeded expectations by miles".

And what about Tekken 7? That game shipped 2.8 mill back in March. Where's the source proving is hasn't crossed 2 mill at retail over 4 months later?

I'd say they're in a better position to allocate resources next year. Keep in mind, EPD still had teams working on 3DS titles prior to the Switch's launch as well, so those developers still needed to finish their projects. Now that those games are done, the staff working on those 3DS titles are now freed up for Switch projects. I still feel that Nintendo can do both ARMS and Mario Kart at the same time, with nothing really stopping them. 

I dunno, it remains to be seen if EPD are indeed capable of providing a steady stream of high caliber Switch tier software. They haven't yet proven that they can without suffering down years as a result.

But curl, how do you explain ARMS' existence in the first place?

After the massive success of Splatoon, Nintendo thought they could create another breakout new IP and simultaneously showcase the Joycon's motion controls.

Ok but then why did it have to be the Mario Kart team to produce it if they should rather make games with higher sales potential?

They massively overestimated its sales potential.