You don't understand the difference between short-term data and a long-term trend. Or, rather, you do, but you are pretending you don't when it suits you.
Broecker's blue line is a long-term trend. The red line is based on observations and is influenced by both the long-term trend and short-term fluctuations. If we had been having the same discussion in 1991 then the red line would be above the blue line and, based on your logic, you'd be bound to say that Broecker
underestimated CO2 sensitivity. But would you have siad that? No, of course not, you'd be arguing that it's just a short-term fluctuation.
When short-term data shows a very large warming in a single year, such as 1991, you'd say it is only short-term observed data (and you'd be right, and I'd say the same thing). When short-term data shows cooling, you say it disproves a long-term trend (and suddenly you are not interested in being right any more). This is why people like you are cherry pickers.
The thing is that people like you can only be convinced if we have a time machine, go to the year 2200 and measure the temperature. You are not interested in scientific process at all, only observed data - and only when it suits you.
I'm a marine geologist, I sometimes reconstruct ocean temperature changes from tens of thousands of years ago. There were no satellites then, or thermometers. So we have to rely on proxies and science. Similarly, our satellites cannot yet tell us future temperature, because we have not invented time machines yet. So we have to rely on science and our understanding of long-term trends.
I mean really, people like you have to be consistent and also not trust weather reports. If the weather report says 13 mm of rain will fall tomorrow, but only 9 mm of rain falls tomorrow, do you never trust a single weather report again and be convinced that Met Eireann has a hidden agenda to overestimate rain? Maybe they are in cahoots with umbrella producers...