Agree completely.

The only exception is if someone wants to deep-dive into a poll to explore methodology and account for it in the analysis. Almost none of the articles we see posted do that though - they tend to state the (real or imagined) results and go outward from there.