I doubt it would be anymore difficult then programming for PS4 and PS4 pro, or a few generations of an idevice, or multiple computer configurations. Although I imagine we may have games that can't go portable (or are just really bad in portable mode).

Developers are already having to do that with the PS4 and PS4 Pro, so I doubt this would be a problem. 

Well you see, developers don't have to upgrade their game for the PS4pro, unless they want to.

If true I'm expecting this to sell for $399. And I don't think that people will be willing to spend that much on a 720p tablet.

Not necessarily too much trouble, many games are designed for PC as well and have lower settings than the Xbone and PS4.  So long as there's not need for a bunch of changes to make it work with the Switch's architecture, reducing some things to make it work better wouldn't be a major issue.  But as I already said, I think Nintendo will actually just shoot for one performance in both configurations that gets them as close to Xbone parity as humanly possible.

Well, Im nowhere near a tech expert, and Im actually quite uninformed in this subject, but with the "closer to the 1.5 TF" could could simple be that it manages to squeeze out 1 or two extra TF in terms of performance to rend the game closer to 1080p or/and at 60 fps instead of the 720p that suposedelly the handheld has.Again, Im supposing that is possible and please correct me if Im simly wrong.But isnt, according to the article, the Switch using full floating points, couldnt it theoretically with a cooling system that handles it, achieve the double of the efficency and getting close to the 2 TF? 

Eh, not necessarily.  I think 299 to 349 would still be well within reach depending on deals made.  But it's hard to predict without the full picture.

I think it really depends on how efficient and easy to port is the Switch, and how good is the tools that Nintendo made(Im supposing there is one) to make both "versions".In that regard, they could simply have made things easy due to the customization of the chip(after all, thats the whole concept of the system).I mean, I heard somewhere that the optimization of PS4 games to PS4Pro was quite easy and not a burden at all.I think it was Tabata that said it, not sure.If thats true, I cant see why it would be the much of a hasstle for the Switch

Not everything can be done in 16fpp, that's why standard GFLOPS and TFLOPS are measured based on 32fpp. Now having good 16fpp performance can be good for those who take the effort to take advantage of it by having those things the can be done in 16fpp handled in that way, but it won't double performance.  The big issues are cost, battery, and heat though.  I just don't see a means by which Nintendo and Nvidia could get anywhere near 1.5 to 2 TFLOPS while keeping the cost down, battery acceptable, and heat in check.  It's much better to take advantage of Nvidia's ability to make maximum use of each GFLOP and shoot for something more reasonable like 1 to 1.1 TFLOPS and use the advantage over AMD's tech to get real world performance parity (or very close) with the Xbone since that could almost certainly be done with the systems price, battery, and cooling targets maintained.  

Not necessarily.NVidia isnt in the best position, especially regarding the console space, so Nintendo could very well bargained a very good deal.I still see it being sold at 300.But I think with this 250 is out of the question.

And you dont get tired of calling the Switch a 720p tablet?Its not that.Well, not only that anyway.