(My starting point in lay discussions of climate)
The main point can be stated very simply. When CO2 is added to the atmosphere, this added CO2 acts like the input signal to an amplifier. Water vapour has no such action. The reason is that water can evaporate and water vapour can condense or freeze, depending on the temperature. The concentration of water vapour is therefore strongly influenced by temperature. This makes it act as a so-called feedback. The feedback is positive. Anything that makes the atmosphere warmer on average tends to make it moister on average, other things being equal, simply because warmer air can hold more water vapour before the vapour condenses. More water vapour means more greenhouse effect, hence still more warming. In other words, water vapour is part of the amplification mechanism.
Regarding the natural range of variation of atmospheric CO2, ca. 180 to 290 ppmv, a good recent description of the care with which CO2 is measured in ice cores can be found in another peer-reviewed journal, Philosophical Transactions of the Royal Society: Wolff, E. W., Greenhouse gases in the Earth system: a palaeoclimate perspective, Phil. Trans. R. Soc. A 369, 2133-2147, © 2011 The Royal Society. This paper also presents some of the measurement data showing the range of variation. See also this useful quick summary on the website of the Cambridge Centre for Climate Science.
Regarding `most of this rise is due to human activity' (taking CO2 from 280 to 390 ppmv), the evidence is strong, though subject to minor uncertainties. The most important feature is the steady rise over the past century or two -- not the transient, short-term fluctuations that are inevitably superposed on it, such as the regular seasonal cycle and other transients impacting the biosphere. These transients include effects from El Niño and volcanic eruptions. The timescales range up to several years, with associated rates of change of atmospheric CO2 that are relatively large, but temporary. The bottom line is that the total amount of fossil fuel burnt and cement produced since pre-industrial times is enough to account for the magnitude of the steady rise, even after allowing for oceanic absorption of CO2 and for the net absorption or emission from the natural and cultivated terrestrial biosphere, after smoothing out the transients. (This net terrestrial contribution appears to have been relatively small to date, though uncertain even as regards its sign.)
Regarding the disturbance to the carbon cycle -- and hence to the climate system -- the effects will be essentially permanent, from a human perspective, unless future generations decide to pull CO2 back out of the atmosphere. The natural carbon cycle has multiple timescales, including timescales far longer than a human lifetime. A readable account is given in a small book by carbon-cycle expert David Archer, The Long Thaw: How Humans Are Changing the Next 100,000 Years of Earth's Climate, ©2009 Princeton University Press. One of the most telling lines of evidence comes from recent work on the PETM (the Paleocene-Eocene thermal maximum, showing in ocean sediments the long-lasting effects of a large natural injection of carbon into the atmosphere that took place about 55-56 million years ago. The climate disturbance lasted for more than 100 millennia, producing severe ocean acidification and mass extinctions.