Thinking is quite a bit faster. I have caught myself altering phrases mid-pronouncing because i was thinking more about it while talking and thought of a better thing to say. In practice that tend to produce garbled words.

You can probably notice similar things yourself if you pay attention, and that's the closest to evidence we can get at the moment due to lack of telepathy or equivalent technology to actually measure these things reliably.

As far as Neuralink goes, assuming it won't be made in the stupidest way possible (reading commands spoken by imaginary voice inside your head), it could be quite a bit faster than hands or speech.

I do this too; honestly it seems natural. Works just fine when typing but can be a little awkward vocally.

It's pretty trivial to prove that thinking is faster than speaking -- just read a 200+ word article out loud vs. "in your head" and the difference is clear. The gap in speed will become larger as the overhead behind speaking grows.

The real question is: does it matter? All around me I see people interrupting each other and failing to comprehend what each other is really trying to say due to impatience. Neuralink is interesting, but would probably exacerbate that particular problem.

We think in images and sound and taste and smell with emotion and memory. We speak with linear language, inflection, and timing. Obviously thought has a higher bandwidth.

If you're asking for scientific evidence I can assure you there is none. It's like asking if a bread box is bigger than an elephant. No, it's not.

But then again people ask some interesting questions.

Yes, yes there is.