Leave it up to the dev. Likely speaking what you'll more likely get are games like Resident Evil 2 Remake that simply cannot run on the OG Switch that are Switch Pro-only ... but really you wouldn't have gotten that game on the OG Switch anyway, so what exactly have you lost? Nothing. If anything you've gained something because that game will be there waiting for you, whenever you feel like you want to upgrade. 

Just like today I can buy a new PC GPU and I know there are new games waiting for me to play that my GPU can't run. And that's great, that's power to the consumer, I choose when I want to move up, it's not dictated to me by a company deadline like I'm a 10 year old that needs to beg my parents to buy me my video games. 

My issue isn't games that wouldn't have run on base Switch anyway. My issue is games that could have run okay on base Switch becoming barely playable 20fps chugfests cos devs prioritize the stronger hardware and put the bare minimum of effort into making the base version run.

There wouldn't be much incentive for a dev to do that if OG Switch users are the majority, which they would be anyway through 2021 most likely. 

Even if there are a handful of devs that want to piss off the consumer base by doing that, that's on them, they lose business. More likely what you will see is the same games that would have come out on Switch, come out on Switch. What will be added are PS4/XB1/even PS5/XB2 tier games that simply are impossible to run on the modern Switch, but those devs would be interested in offering versions of some of those games (ie: Kingdom Hearts 3) for some extra sales in the Switch ecosystem. 

Long term IMO when people get used to the new setup they will realize they probably like the new setup better ... driving a hardware into the ground until its badly outdated and losing brand momentum is not a positive. It's not even good for business, it was a reality of the hardware model of the past because largely in the past the majority of gamers were kids that had to rely on parents to buy their game hardware for them so you had no choice but to space things out. 

But today? We live in the Apple world and consumer expectations are dramatically different, not only is iterative business models OK by consumers, most actually *prefer* that model. They want the company to give them multiple choices and then they choose when they want their next iPhone, Samsung Galaxy, Playstation, XBox, Switch, Apple Watch, iPad, whatever. It's up to the consumer to decide. 

Then Nintendo are shit out of luck regardless ? The console cycle just simply needs a hard reset irrespective of the next generation system featuring BC or not and that's the way it will remain because if their only way of maintaining higher hardware sales is through selling new iterations of their franchises advertising the capabilities of their next generation system then it won't mean much whether or not they adopt your so called "PC model" since they only did it in name only ... 

After all, aside from releasing a more expensive model to appease the 3rd party developers what are the other advantages your model offers compared to existing home consoles who also offer upgraded systems as well ? Should Nintendo titles also adopt a continuously updating development baseline or should they also stick to a hard baseline target and change it as if there was a new upcoming generation just like how every other console manufacturer did so far ? 

Most devs are generally cheap and lazy and will take the path of least resistance; if that means stronger hardware making development easier than optimizing something properly on the base Switch, they'll do that, and we'll have to suffer for it.

Take a game like Wolfenstein II for example; in its current state it holds a stable 30fps most of the time on Switch, but I'm willing to bet if a stronger model was available it would run at 30fps on that and struggle along at 15-20fps on the base model.

Lazy developers also had a tendency to get punished with lower sales. But sure that can happen, there are pros and cons to that approach, but from a business POV, I think there are more pros than cons.

50% of Nintendo's hardware transitions have been failures/dissapointments, you can't bank the entire company on a traditional Switch 2 going swimmingly well. What's your back up plan if it doesn't? The better plan is to not put your company in that position to begin with and change to something that's far less risky of a hardware model. PC has been doing something else for decades and its worked fine there, Apple is making more money than all three console makers with their model. 

There are pro-consumer arguments for an Apple/Steam-PC like model too, I think the freedom of being able to upgrade whenever you want and having games available right from day 1 instead of having to wait months/years for the library to build up is a significant pro. 

They can maintain hardware momentum with hardware revisions like they did with the DS. It doesn't need to be a full-on Switch 1.5 with a big power boost, it could just be a Switch Deluxe with a bigger and higher quality screen. That way you avoid consumer interest waning until it's time for a proper successor.

The DS era was a different time, people were easier to impress, the full weight of what Apple was about to do and the smartphone industry had not really shaped people's expectations. 

So something like the OG DS to DS Lite was considered an impressive change for 2006, but that's 12 years ago.

The 3DS revisions ... every single one of them basically resulted in no sales boost, and in many cases it actually resulted in a decline. 

Past that, I don't even think the 5-6 year cycle works for Nintendo, it was a set up born out of the 1980s when Nintendo had 100% third party support and could easily have 5 years of content. 

In the post Playstation era where Nintendo doesn't get that anymore and has to rely on basically 7 or 8 main franchises to sell a platform ... it's hard to not be tapped out of content even by year 3. The Switch for example will have used up basically all the A/B-tier Nintendo franchises by the end of 2019 (3D Mario, 2D Mario, Zelda, Pokemon, PLG, Mario Kart, Splatoon, Animal Crossing, Fire Emblem). It's not a coincidence most Nintendo hardware sales really start to decline by year 4/5 of their hardware cycles, because that hardware cycle was never designed to be a setup where the system was mainly sold by Nintendo games. The NES and SNES were not like that. 

It's very hard to maintain 5-6 years of 15 million+ hardware sales (which is likely what Nintendo wants) when you're basically relying on just one company's franchises and then some secondary/tertiary support from developers. 

Switch is not the 3DS, it's selling as well at $300 as 3DS did at a $170. There's a lot of interest in various hardware revisions of the original Switch.

As for "they won't have any more system sellers after year 3", even if that were true, a Switch 1.5 in 2020 and every 2 years thereafter doesn't solve that anyway.

Will be in the same switch like console.
The same screen size, but with smaller bezels
Improvements on joycon. each side has 1 handle, but can open to 2 handles when using a single joycon, and would look like a normal controller. (id have to draw, because is difficult to explain...)
In the technology, They will keep the tegra line with better chips on smaller dies.
1080p display on the screen, but the dock will feature the DLSS of NVIDIA (that is available now on RTX 20xx cards), because they can improve the resolution post rendering using AI. So, 1080 on HH, 4k with AI on dock.
Full BC.

dock would be more expensive, though. So it will be launched in 2 skus, one with dock, and other without it (although, it would have a mini hdmi port to output the 1080p image to screen even with no dock). 400$ for the switch 2 4K and 250$ for switch 2 with no dock. 2024.