"A recent thread here had various people recalling their memories of a
vintage  Sky & Telescope article about reflectors of different wavefront
quality.  Given that the memories differed--and that I had just been
browsing that old article--I thought I post a summary.  I note some
interesting wrinkles in the testing.

The article by Ceravolo, Dickinson, and George appeared in the March 1992
issue, page 253+.  The purpose was to see how well observers could
distinguish--in a blind test--among reflectors of known, varying quality. 
They used 6" f/8 mirrors.  

(Is there any reason to think that the results might differ at lower f
ratios or in larger mirrors?  I ask because we should figure out how well
the results generalize.)

First, the primary mirrors were near 1, 1/2-, 1/4-, and 1/10-wave
peak-to-valley (PV).  (The mirrors were ground, figured and polished by
Ceravolo and tested using a Zygo interferometer.)  

Second, the error that was put in was spherical aberration. Could this
affect the generalizability of the results?  Might other types of
error--turned edge, rough surface, etc.--have been more or less apparent?

Third, it should be noted that commercial secondary mirrors of varying
quality were used!  Two were 1/20 wave, one was 1/10, and the last was
1/2.  They put the three best diagonals with the three best mirrors.  (So
much for a standardized test of the primary's quality.  Although given the
pairings, and the high quality of the 3 diagonals, it shouldn't matter--or
should it?)  

The blind testing was done separately by Terrence Dickinson and Douglas
George, and then at Stellafane.

FINDINGS

DICKINSON

Dickinson did his tests over five separate evenings from his rural home. 
He reported on observations of the globular M15, the Orion Nebula, NGC
2158 (cluster beside M35), and Zeta Orionis (2nd, 4th magnitude double
star with separation of 2.4").  An 80mm Celestron refractor matched the
1/2-wave 6" scope on Martian detail but gave a more pleasing image, and
beat it on the double star.

He found that the 1/4 and 1/10-wave mirrors gave generally comparable
views, although on nights of good seeing, he did spot sharper and finer
detail in Jupiter's belts and subtle differences on Mars with the
1/10-wave mirror.  Both  gave detail on Mars that approached a 5 1/2"
Astro-physics refractor.

"Frankly, I am surprised by the results. . . On those rare nights of
exceptional seeing, maybe the 1/10-wavefront mirror would show its stuff,
but even then I doubt the difference would be noticed by most observers,
except in rigorous side-by-side comparisons on selected test objects"
(255).

He commented that a 12 or 14" reflector need be only 1/4-wave to give peak
performance (256).  

GEORGE

George, too, found the 1/4- and 1/10- mirrors were comparable, save under
"careful scrutiny under excellent seeing" (257).  He looked at the Orion
Nebula, the globular M79, and Jupiter (including a shadow transit).

He was "amazed" that a well-made 6" approached the performance of a
quality 7-inch StarFire refractor.  But he also felt that "many amateurs
would be quite happy"  with the performance of the 1/2 wave scope.  (Near
the end, he states this as "most amateurs would be satisfied".)  Still, he
noted that 1/4-wave is "definitely a minimum standard" for planetary
viewing and that the diffraction limit of 1/4-wave is a "myth" because the
1/10-wave mirror did produce a better view.  (In this, he is diverging
from Dickinson, seeming to suggest that we should seek better wavefronts) 

He found that the focus "snap test" readily distinguished the 1/4 and 1/2
wave mirrors (meaning that with the better mirror, the image snapped into
a crisp focus rather than being mushy and never quite focusing sharply). 
He also noted that the star test--under very good seeing--could separate
the 1/4 and 1/10 mirrors.  (Although in the body of his report, he did not
find such clear distinctions using the star test for the 1/4 and 1/10
mirrors.)

STELLAFANE

Alas, the testing done at Stellafane was only of Polaris!  Alan MacRobert
conducted the testing and gathered reports from 103 observers. S & T only
used the 3 best scopes--that is, the 1/2, 1/4, and 1/10 wave ones.  

No matter what one's level of experience, nearly everybody could pick out
the bad apple in this bunch (1/2-wave).  The 1/4 and 1/10 comparison
proved more difficult.  

"Nevertheless, about two-thirds did correctly identify the best scope as
giving the best image.  This was despite the fact that the atmospheric
seeing was generally average to below average"  (256)

But, get this.  Many viewers did *star tests* to make their
judgments--rather than relying on the views themselves.  (Given the
purpose of the test, this seems to be cheating!)  They found it easy--that
way--to distinguish between the two best mirrors.  Interestingly, the star
test was effective even though the seeing conditions were only average.  

Relying, however, just on the in-focus views to make the discrimination
was much more difficult.   Still, more than half picked the right one
(i.e., which was better than random guessing).  Some figured out which was
best by seeing how well the 9th magnitude companion showed up.

(This again suggests that the views were generally similar in the 1/4 and
1/10 scopes.  A possible wrinkle here, however, is that maybe those who
relied on the view rather than the star test were less experienced. . .)

MacRobert reported that the seeing conditions affected people's
confidence--when the skies steadied, there was more confident selection of
the 1/10 wave mirror as the best.  

The conclusion:  "when the seeing turns good, many persons can indeed make
out a slight difference in sharpness and clarity between 1/4-wavefront and
1/10-wavefront telescopes in focus."

Note the phrases "good" seeing and "slight difference".  But again, this
was only of Polaris.

I highly recommend the article.  The details and descriptions are
informative and make for  fascinating reading.  You can often find old Sky
& Telescope issues at a local university or public library."