Sure, if you're listening to people like David Suzuki you'd have that impression. But on the other side there are people like Chris Landsea - here was his resignation letter from the IPCC:
http://cstpr.colorado.edu/prometheus/archives/science_policy...
The IPCC's charter is to tell us about risks related to climate change. Not the benefits. If there weren't any risks or we were fully informed about them, the IPCC would have no institutional reason to continue to exist. So naturally it focuses on the latest big scary "we just noticed THIS risk!" stories and puts much less emphasis on "it turns out we were wrong about THAT risk!" stories. For all we know, each new report could be exactly the same as the last one in terms of the net overall danger documented and it would still look like things were "getting worse", because areas where things are "getting better" generally aren't mentioned or are soft-pedaled.
For instance, the IPCC once tried to claim a high certainty that there'd be more hurricanes in the future due to climate change (see Landsea's letter linked above), but now they either don't make such claims or assign them a much lower certainty level. The IPCC once claimed Himalayan glaciers would be gone by 2035 ( http://www.theguardian.com/environment/2010/jan/20/ipcc-hima... ); that turned out to be a mistaken claim based on grey literature. If there were some sort of a rundown in each report listing all the ways things "are worse" AND all the ways things "are better" since the last one it'd be easy to keep score. But there isn't, and the summaries and press releases emphasize any mentioned "this is worse" stories because bad news travels fast.
Suzuki complains that the IPCC report doesn't focus on the loss of arctic sea ice, but if they DID mention sea ice, they might have to mention that overall sea ice levels worldwide are currently above the long-term (30 year) average because we've gained more sea ice cover in the antarctic than we've lost in the arctic. And so on. (Given a big, complicated planet you can always find SOME areas or trends that seem to be "getting worse" but that doesn't mean throwing them - and only them - into the mix would make the report more accurate.)