Would already be too low in my opinion to be honest.
What I not only expect but demand from a switch two is to deliver the same performance or more than the rumored XBox Lockhart when docked,
and around halve undocked.
HOW it is done doesnt matter much to me, as long AS it is done.
This could also mean, that the hardware inside the switch two alone is not capable of doing so, but combined with suplementary hardware inside the dock
it would work. (Which would make the Switch 2 an actuall hybrid btw, unlike Switch 1)
With Lockhart, Nintendo has the unique chance of getting full 3rd party support (not just indies and medium japanese games)
despite having compearativly much weaker hardware. And Nintendo would be stupid to pass up on this upportunity.
I'd think it would take them another 1-1.5 years to get to be able to get that same 4 tflop as Lockhart for $300-350 in a handheld form factor (so 2024 or early 2025), (...)
Not necessarily. It's just a question of how you build the product.
In my opinion, the easiest way to get to that level is by having the portable only use hardware which can "produce" 40~50% of Lockhart's performance.
Aka only the hardware that is meant to be used on the go.
However, the dock itself contains a supplementary CPU/GPU/other which combined with the chip present inside the portable unit amounts to Lockhart specs.
Remember, not everything HAS to be carried around in the portable unit.
Just leave in the dock what you dont need in portable mode.
Think of it as a workplace setup. You have your notebook that you can carry around with you, when you work at your desk, you connect a mouse, keybord and
screen to it. In this case however, the mouse, keyboard and screen would be a second gpu/cpu/other.