This article is headed:
“EARTH was a BAKING LIFELESS DESERT for 5 MILLION years”
OK, thinks I, are we talking about a couple billion years ago, where there is no evidence of life of any kind? Or is there really a clean gap in the history of life on earth, where life covered the globe for millions of years, vanished, and returned? Nope. Turns out that after the Permian die-off, earth was still what we would typically call ‘teeming’ with life – just not in the tropics, and just not with big forms of life, like fish and trees.
The problem here, at least potentially, is that someone might be lead by this headline to assume that all life was wiped out and then re-emerged 5 million years later, which would be something. Just not something there’s any evidence for. A sufficiently dogmatic and uninformed person might even imagine that this total-destruction-of-life, reemergence-of-life thing proves that life arose – and arises – inevitably out of completely natural processes. which may be true, but is not proven or demonstrated by the Permian die-off.
A more biologically sophisticated reader might agree with the headline to the extent that today’s ‘lifeless, barren’ deserts are, you know, often teeming with life – just not the kind (and size) of life you’d find in a forest or ocean reef. Death Valley blooms with wildflowers; brine shrimp show up on salt flats.
But the article is really only talking about the tropics, leaving the temperate and polar zones to ‘teem’, as it were.
This is a minor and relatively harmless case of lying via science headline, yet it contributes to the whole science-as-belief-system error so damaging to the modern world.
Yard Sale of the Mind 
Stromatolites of fossilized oxygen-producing cyanobacteria have been found that date back possibly to 3.5 billion years ago but clock in definitely by 2.8 billion years ago.