CO2 is only a tiny fraction of the atmosphere - just shy of 400 parts per million (ppm) right now. But that is some way above 'normal'. Atmospheric CO2 concentrations have varied massively over the history of the planet, but in times relevant to humanity (e.g. the last few hundred thousand years, say), the amount of CO2 in the atmosphere has stayed fairly low, around the 280 ppm mark. The effect that this increase has on global temperatures is not entirely known, but the fact that CO2 levels are alot higher than they were just 200 years ago cannot be explained by natural mechanisms alone (and can be more readily explained when human activity is taken into account).
Your other two points above highlight for me what the danger is in making such changes to the ecosystem without any regard for the consequences (although I don't mean to sound like this has been done on purpose...). I fully agree with your Goldilocks analogy - it has to be just right... all the more reason then to be extremely cautious about making changes, either accidentally or on purpose. We've already done the 'accidentally' bit - unfortunately, many proposed solutions involve 'on purpose' changes, which could (and probably would) be just as dangerous.
Every living thing on Earth today owes its existence to the fact that the Earth's ecosystem has provided a suitable environment for them to survive and thrive. Of course, natural climate change will always ensure that species come and go as their environments change accordingly. But that we don't fully understand how the climate system responds to any particular forcing (either natural or manmade) is not a good reason to assume that everything will be OK - if anything, it is a good reason to be extremely cautious (or at the very least aware/mindful) of the potential impact of tinkering with a system that we utterly depend on. Climate change always has and always will happen, but the danger is that we do (and most likely already have done) something that provokes a climatic response on a timescale that our society (and alot of animal life) might struggle to cope with.
I agree with a lot of what you say, but I think there's this mindset that the climate is this fragile thing. Maybe it is, but I find it unlikely that the climate has remained relatively liveable for however many millions of years if it's really on a knife edge. As far as I know there have been eruptions and the like in the past that have at least for short periods driven particulates and CO2 far beyond anything humans have managed.
And this is sort of the thing. If the climate responded directly to CO2, we'd all be boned already. We don't know what the ecosystem is capable of dealing with, long term or short term. It's wise to play it safe if possible, but at the same time there's no real reason to think that we're in a doomsday scenario.
I'd refer again to the recycling boom. When that was in full swing, the press would tell you that if we didn't all start recycling the planet would be unliveable by our children's generation. That turned out to be a load of twaddle, and actually all that was needed was some moderate effort to live more sustainably with respect to resources.
I see the climate change movement as the same thing. It's not a doomsday scenario, it's simply something that we should incorporate into our lives. Be mindful of how you use power. Consider whether you really need to buy new goods, or if you can get used. Think about your transport options. Etc.
Re: modern CO2 staying below ~280ppm. That's more or less true of the time period you say, the last few hundred thousand years. But there's been similar lifeforms on earth for millions of years and the CO2 has potentially been as high as several thousand ppm, although obviously the margin for error is fairly high on measurements that old. The models run a little lower, but still in the 1-2 thousand ppm range at peak.
http://en.wikipedia.org/wiki/File:Phanerozoic_Carbon_Dioxide.png
That's the best image of it, but the data is drawn from:
http://droyer.web.wesleyan.edu/PhanCO2(GCA).pdf
I'm ignoring the very early days where the earth was likely unlivable for most of the species extant today. But about 300 million years ago it wasn't so very different to today, and the variation between then and now is enormous.