As you said, except for NSMBU those were 2018 releases. And except for NSMBU and Mario Maker 2, all the high-profile Switch releases for 2019 were after the Lite was released in September.

Nobody has yet to satisfactorily prove that any one game (or four closely-released ones) can improve their system's baseline for many months on end. We've never really seen it with other systems before, so why assume it's happening now? Also, having good lifetime sales doesn't make a game a system-seller. Sure, Smash was an undeniable blockbuster and Pokemon Let's Go did very well, and there's solid evidence that both moved some surplus hardware. But neither NSMBU nor Mario Party appears to have moved the needle for software in the short term if what they did in Japan is any indication. Their initial sales in Japan weren't all that impressive either despite their solid lifetime sales; according to Media Create, first week sales for NSMBU and Mario Party were 166k and 142k respectively, relatively mediocre starts compared to a lot other popular games. In the U.S., NSMBU had already dropped out of the Monthly Top 10 in February and the YTD Top 10 by April, and Mario Party dropped out of the monthly Top 10 after the 2018 holiday season (Pokemon Let's Go's legs are hard to determine as NPD logged the Pikachu and Eevee versions separately).

You could make a case for January 2019 feeling some residual effect from Smash as there have been a few system-sellers to still have a small effect in the month after their release, but that's about it. Except for maybe Splatoon (and only in Japan), there's never been a single recorded case of any obvious increases in hardware sales two months after a notable system-seller's release. The effects of individual games on hardware sales are often grossly overestimated. Their effects are very short-term. And again, we don't see continued strong lineups over time result in growth in baseline sales for other systems, so why assume the Switch is somehow unique in that regard?

And FWIW, we didn't see a consistent increase in the Switch's baseline in Japan like we did in the U.S. While sales were up in January & February 2019, by mid-March it was down to being about flat from 2018 and it stayed like that until the Lite was released.

Different situations entirely.

Since your premise is irrelevant, your conclusion based on it is invalid.

As you said, except for NSMBU those were 2018 releases. And except for NSMBU and Mario Maker 2, all the high-profile Switch releases for 2019 were after the Lite was released in September.

Nobody has yet to satisfactorily prove that any one game (or four closely-released ones) can improve their system's baseline for many months on end. We've never really seen it with other systems before, so why assume it's happening now? Also, having good lifetime sales doesn't make a game a system-seller. Sure, Smash was an undeniable blockbuster and Pokemon Let's Go did very well, and there's solid evidence that both moved some surplus hardware. But neither NSMBU nor Mario Party appears to have moved the needle for software in the short term if what they did in Japan is any indication. Their initial sales in Japan weren't all that impressive either despite their solid lifetime sales; according to Media Create, first week sales for NSMBU and Mario Party were 166k and 142k respectively, relatively mediocre starts compared to a lot other popular games. In the U.S., NSMBU had already dropped out of the Monthly Top 10 in February and the YTD Top 10 by April, and Mario Party dropped out of the monthly Top 10 after the 2018 holiday season (Pokemon Let's Go's legs are hard to determine as NPD logged the Pikachu and Eevee versions separately).

You could make a case for January 2019 feeling some residual effect from Smash as there have been a few system-sellers to still have a small effect in the month after their release, but that's about it. Except for maybe Splatoon (and only in Japan), there's never been a single recorded case of any obvious increases in hardware sales two months after a notable system-seller's release. The effects of individual games on hardware sales are often grossly overestimated. Their effects are very short-term. And again, we don't see continued strong lineups over time result in growth in baseline sales for other systems, so why assume the Switch is somehow unique in that regard?

And FWIW, we didn't see a consistent increase in the Switch's baseline in Japan like we did in the U.S. While sales were up in January & February 2019, by mid-March it was down to being about flat from 2018 and it stayed like that until the Lite was released.

And no individual game, nor any particular lineup of notable games released over a brief period of time, has ever caused an increase in baseline sales lasting for months on end. It. Just. Doesn't. Happen. A system having a good performance in general does depend greatly on the strength of its library, but actual sales growth for periods longer than a month or two is, as far as can be determined from the data, not the result of any one game or even a group of closely-released games. We've never seen a strong lineup in one year (or half-year, or quarter) do it with any other system before. Why assume it's suddenly a thing with the Switch?

From my experience, once you're locked with your family, all of a sudden access to a TV or console becomes a problem. Switch is a perfect solution. It's personal, convenient, for everyone and cheaper than buying a new gaming PC/laptop or PS4 + TV/Monitor. For the price of a single decent gaming laptop, you could buy 4-6 Switches for the whole family. And even if that's not an issue and the whole household would like to play something together, Nintendo is known for its local co-op/split-screen family-friendly games. Not to mention fitness games. It's just a perfect lockdown system. 

Re-read what I said about Mario Party and NSMBU. You'll see that I said they didn't have good initial sales. Their lifetime sales are good, not because they started off well, but because they managed to have legs just good enough to generate a good lifetime tally. NSMBU in particular has been quite leggy; while it shipped only 4.1M by June 30, 2019 and 5.85M by Dec. 31, 2019, it's now at almost 10M. Not a record-smashing start (esp. in Japan), but good enough long-term sales to do what it did.

Regarding system-sellers. There are very few games that cause any significant increases in hardware sales. That's a simple fact. The vast majority games do precisely jack squat. Even some games that are very popular don't move any measurable amount of hardware. It's even harder to notice any effects when sales are tracked monthly like they are in the U.S. (a game that causes a 50% bump in weekly sales in the week of its release would only cause a 12.5% increase for a 4-week month, something that could be easily overlooked as statistical noise). Sales are tracked weekly in Japan so we notice more system-sellers, but the vast majority of those do relatively small one-week bumps. Some have a residual effect lasting a few weeks, which is consistent with what we see in the U.S. with most clear system-sellers having an effect in just their release month, with a select few having some sort of residual effect in the next month.

And no individual game, nor any particular lineup of notable games released over a brief period of time, has ever caused an increase in baseline sales lasting for months on end. It. Just. Doesn't. Happen. A system having a good performance in general does depend greatly on the strength of its library, but actual sales growth for periods longer than a month or two is, as far as can be determined from the data, not the result of any one game or even a group of closely-released games. We've never seen a strong lineup in one year (or half-year, or quarter) do it with any other system before. Why assume it's suddenly a thing with the Switch?