OK, you admit this: 在起步的瞬间离合器打滑是不可避免的

(1) Now if you want to keep engine revving at 6000 rpm from the rest (起步), all the way up to 60 mph, do you agree that 离合器打滑 is no longer just a 瞬间? That is, I claim that your best CVT has to have a finite low ratio. If the lowest ratio is similar to the 1st gear on a conventional manual, you'll be slipping the clutch for several seconds if you keep the engine at 6000 rpm. Do you not agree with this?

(2) As long as we are slipping the clutch for a while, let me ask you this: during this time, is the clutch transmitting 100% of the power it is normally capable of when it is closed? The answer is NO, and it comes from elementary physics. If it is a real clutch with two discs, then in closed state it transfers force via stationary friction force, and the coefficient (to be multiplied by pressure) is higher than the coefficient when it is slipping (not to mention the pressure is higher too). You just can't get around this, ideal tranny or not, do you agree?

(3) Let's say we pick a definite number, that in a slipping state, the clutch can transfer at most 90% of what it normally can. Furthermore, if you accept that hp comes in a curve, there is a point before 6000, say at 5000, where the engine makes 95% of the peak hp, no? Do you agree with this assumption so far?

(4) Now let's review the 0-60 process. You want to hold rev at 6000, and because of the finite lowest ratio, up to say 30 mph (your example) the clutch will slip, and from then on you can close the clutch and let the CVT do the work. But let's look at when the car gets to 25 mph. If you closed the clutch at this point, the rpm will drop to 5000, you'll only have 95% of the peak power, but you transmit 100% of that instead of 90% of the peak, so you are better off. Do you not agree with this?

(5) To recap, your strategy of holding 6000 rpm all the way cannot be optimal for 0-60 runs. If you started off at 6000 from the rest, it is always better to dip the rpm and close the clutch early. Exactly how early depends on the hp curve and clutch design. But there is a dip. Do you not agree with this?

(6) As long as you give me a dip in rpm during the 0-60 run, I can make an engine with the same peak hp and out-performs your engine, just by bumping the hp curve in the range of the rpm dip slightly, while not exceeding the peak hp. You have to agree with this as well, no?

(7) Now that I have some extra performance to spare, I can certainly waste a little by having a slightly lower hp curve overall. Therefore I get an engine with lower peak hp than yours, and performs better for 0-60. No problem here, no?

To conclude, as soon as you admit that a slipping pair of discs cannot transmit force to the same efficiency as a closed pair, I've proven that peak-hp alone doesn't translate into faster acceleration times from the rest.

Instead of saying something general again, if you don't agree with my argument, why don't you point out exactly which step you disagree with, and then we'll continue from there.