The test shoots in this article illustrate that the ANSI lumen spec and the new Color Light Output spec can both be misleading, but for different reasons.
The ANSI lumen spec is often misleading because it does not take either color brightness or color accuracy into account. Knowing that a projector is rated at, say, 3000 ANSI lumens does not tell us anything about how it will perform with full color images. The picture's whites may be bright, but the colors may be dim, or poorly saturated or wildly inaccurate. The picture may have an objectionable tint. Most commercial/business class projectors (regardless of the technology used) have flaws of this nature that the ANSI lumen ratings ignore. Once you adjust the picture controls to get rid of the color flaws, a projector can end up putting out much less light than its specs would indicate.
The new Color Light Output spec suffers from similar limitations: it does not take color accuracy into account, and it reveals nothing about whether the projector is calibrated to retain highlight detail. So a projector may have two specs that say "ANSI Lumens: 3000 and Color Light Output: 3000" and still end up giving you a picture you would not want to watch. Once the projector is calibrated for best picture quality, it may put out far less light than its ANSI lumen / CLO specs would indicate.
The Color Light Output spec does point to a real difference between three-chip projectors vs. single-chip DLP projectors. These two different technologies often do produce pictures that contain quite different amounts of color information. When a 3LCD projector and a DLP projector with a white filter have the same ANSI lumen rating, the DLP may produce noticeably dimmer full color images. To that degree, the "white only" ANSI lumen spec can be viewed as misleading and biased in favor of single-chip DLP projectors. In theory, the CLO spec is intended to remedy this bias.
Those who do not support the CLO spec say that its only real purpose is to point out a particular weakness in DLP technology without drawing attention to related flaws in 3LCD technology--most commonly, the unattractive blue/greenish tint you get when the 3LCD projector is in its brightest operating modes. And from a practical perspective, since most vendors will not publish CLO specs, they cannot be used as points of comparison in the same way that ANSI lumen numbers typically are. So in the end, detractors insist that CLO should be seen for what it is--a marketing tool used by 3LCD promoters rather than an official spec that has been accepted and embraced by the industry.
Nevertheless, the fundamental point that CLO makes is valid: the traditional ANSI lumen spec utterly fails to provide buyers with an accurate apples-to-apples comparison of how bright projectors really are when they are used to display full color images.