Some times, they don't know much at all. I have been in disagreement with
many magazines before... I have also seen very large groups in disagreement
with magazines.

I remember reading in a magazine that compared the Ford Mustang to the
Pontiac GTO. The GTO beat the Mustang in EVERY test. it was said in the
article that the GTO posted better times for every test, they even had the
numbers side by side... want to know who won? yeah, the Mustang. how does
that work?

Randall,
  UK magazines like to find faults even with top-notch kit in order to give
unbiased reports.

For Peter's reference, I would make these points:-
This is what the August edition of Digital Camera Magazine (a UK mag) says
about the 20D.

"The Canon EOS 20D is a well built, comprehensively featured SLR that
performs almost faultlessly. We have no criticism of any consequence.
Superb."
The 20D is the magazine's top Best Buy in the DSLR category, and that
category includes the astonishing 1DS MkII, which they felt was overpriced.

I own a 20D, and it *is* superb. My observations on the points made by the
magazine Peter read are:-

-The camera does not over or under expose - it averages to 18% within the
selected metering method, like every camera. Any perceived under-exposure is
most likely a misunderstanding of the metering used. I have not noticed
exposure problems other than of my own making.
-The in-camera b&w filters are OK but not jaw-dropping - shoot colour and
post-process in PS if you want digital b&w shots. Personally I shoot film
for b&w.
-The camera autofocusses on the nearest object covered by one of the 9 focus
points, not necessarily on your chosen subject - that is *why* the toggle
selector has been provided to override this. You get used to this v.
quickly. I have no complaints with focussing accuracy.
-The mirror is noisy - but it is also damned fast. Black-out time is not a
problem with this camera, and you get 5 frames p.s. It is not as noisy as my
antique Pentax SP1000 (ker-LUNK).
-There is no true spot meter, but it does have a 9% spot meter, which I have
always found adequate. For critical shots, you are better off bracketing
anyway.
-The amount of in-camera image processing is deliberately minimised - the
images are *meant* to be post-processed in order to get the very best
results.

Muzz.
A "stupid Limey".

Also remember that many magazines review cameras based on what they think
the average user will experience.
Many labs will compensate for underexposed pictures when printing, thus
making that underexposure less visible.
If a digicam doesn't do so in-camera such reviews will mention that the
camera underexposes shots (when in fact it's possibly operator error which
is now not masked).

All AF cameras do that once in a while, I see no reason the 20D shouldn't
(or why it should be specifically mentioned).

So a higher resolution chip (compared to what anyway?) is now a liability
instead of a bonus?

If you turn on all kinds of in camera corrections of course filters become
ineffective.
Just as they become ineffective (or less effective) when you send your
pictures to be printed in a one-hour lab that thinks there's a weird color
cast and decides to remove it.

I'm not a Canon fan by any means (as becomes very clear looking in my
camerabags), but this review seems written by someone who has some serious
lack of knowledge about the process after you press the shutter.