Two relatively extreme examples.

Digression from notes

Consider Bryan Cantrill on Jails and Solaris Zones | Papers We Love (2016-02-11) – 105 minutes of excitable, fast-talking discussion around a 44-slide presentation.

How fast? Commentary under https://www.youtube.com/watch?v=hgN8pCMLI2U includes a wish for the speaker slow down a bit; the speaker jokes that "To slow down is to admit defeat". There are neither subtitles nor a transcript (I might joke that the pace of discussion defeats the will of most potential transcribers); if spoken at a more 'traditional' pace, I reckon that the discussion could have spanned three hours or more.

Is it good to read notes during a presentation?

At one point, Cantrill appears to rein himself him – one of the digressions (from FreeBSD jails) ends with a salutory mention of Robben Island and whilst presenter's notes are at the rostrum, the reining involved no visible use of those notes. Instead, there's:

• visible presenter focus on the audience's view of the presentation.

At all other times, the audience enjoys presenter focus on the audience; on his maintaining their engagement with what's discussed.

Attention to notes

DCTLib observedDCTLib observed:

… at a premier conference. The introduction was ultra-carefully crafted to set the work into a very precise context, in a well-understandable way, in very short time, while speaking clearly and slowly at the same time. After that, he continued with freely talking about the actual novel results in his talk. The introduction was actually great and helped researchers from related disciplines a lot to grasp the main ideas.

Back to the opening question:

Is it good to read notes during a presentation?

For an introduction of the type described by DCTLib: yes – it can be very good to do so.