Sure, but there's a TON of out-of-band data in voice that cannot be easily replicated in text, that's a fact.

Voice tone, pitch, pauses, irony, sarcasm (both easier to misunderstand in text), etc.

And in person or voice chat have the highest signal-to-noise additional data stream that is body language. "The mouth says yes, but the eyes say no."

Yes, good luck explaining things to people who can't understand basic instructions via text. You might use it for "records" to protect yourself, but getting things done is easier. When I ask someone 'do you understand', in text it might sound very aggressive and sometimes offensive. Via phone or video, it is much more relaxed. If you are concerned about recording, just use screen capture and save it somewhere for a while. Even text cannot be used as a clear evidence, the other person simply says 'I didn't understand that sentence that way'.