While often called a gamut, YCbCr (often abbreviated YCC) is actually a method of encoding pixel information from an image. It would appropriately be referred to as a color model. Originally invented to transmit TV signals in a form that was backwards compatible with B&W TV's during the transition to Color TV's, the YCbCr encoding is capable of storing a full-detail grayscale image (the Y channel) independently from the color channels (Cb and Cr).
The color channels are encoded such that they include a minimum of information, thereby using bandwidth as effectively as possible. The two color channels, Cb (chrominance blue/yellow) and Cr (chrominance red/green) are half-resolution, and containing only two colors each are easier to compress than a single full-color channel. Current variants of this encoding include YCbCr and YUV.
While a rather dated entrant in the modern collection of color models, there are modern uses and enhancements to it. It is presumed that Canon makes use of a high-precision variant of Y`CbCr in their M-RAW and S-RAW image formats. The encoding is used to efficiently encode full detail pixels from a digital sensor at half or one quarter the native resolution of the sensor, thereby making non-redundant use of the bayer color information, while producing a more accurate luminance channel. While not a true "RAW" format, M/S-RAW can theoretically produce images with much richer and more accurate color reproduction, possibly on par with Foveon layered sensors.
Additionally, a modern enhancement to YCC has been created to support color encoding at a much higher level of precision and scope. Where as the original YCC encoding was created to save space and support compression, xvYCC aims to expand the range of colors (gamut) that may be encoded with YCC. Previous YCC encoding was only capable of encoding the rather narrow gamut of sRGB. xvYCC supports a gamut nearly double the size, ideal for modern electronics that have far greater capabilities.