Is it possible to have dialogues with npcs without the need of using audio files AND have the text displayed for some time so people can read it actually? PLEASE do not tell me we have to record blank dummy files to make the text display longer like it was in Oblivion.

That's how it works by default. If there is no audio file, the text displays for quite awhile - you can click on it to advance when done reading, though.

I'm working with quite a lot of dialogue right now, a lot of which uses up most/all of the 150 max character length, and it often doesn't stay on the screen long enough to read through it leisurely while it is unvoiced. I am unsure if it always does that, however, but it does it often enough that I would need to do something about it if I wanted to release the mod unvoiced or even just play it myself as it makes some of the dialogue useless. Because I'm adding voice files I haven't paid enough attention to the issue to see if there is some common thread. But, yes, I can see why recording/using a blank audio file could solve that problem.

At least they fixed that.
Could it be the default Greeting message playing some preset voiced audio? (which is very short).

I guess I really haven't been paying attention - the text displays for about 6 seconds regardless if it's one character or 150 chars. On short ones it seems like an eternity

Gecko did this by writing silent mp3s for all your unspoken dialogue with just a button press. Not sure if there is a FNV version at this point, but you can just make your own silent mp3s with whatever duration you like, and add them to the NPCs dialogue.