About the only "data" you see on this group is "It seems to work for me." I think this is frequently worse than no data since it implies that something is actually a workable technique when infact we have no idea what kind of test, if any, was performed and with what level of skill, or with what favorable conditions which might be different in another case.
Of course every thing I say is just my opinion including the above. (My license to see and speak the absolute TRUTH was revoked ages ago :-). Even if I make measurements or do calculations it is still just my opinion since I might have made a mistake somewhere. My opinion is that if you need more than three points of support on the back of your mirror, then a conventional lever-based flotation cell is the proven, time honoured, way to go. It is simple to construct, intrinsicly self-correcting, does not exibhit any of the deformations discussed below, and is unaffected by surface irregularities on back of the mirror, uneven heating, part warpage, tilt, dew, phase of the moon, etc. Of course my opinion is subject to revision if someone shows me a design that looks better. Here is a more detailed description of my line of thinking.
There are (at least) three kinds of deformations we are talking about. Just discussing the pads/blobs supporting the rear surface of the mirror we have: 1, compression of pads/blobs in the mount (assuming only three pads) a) even compression causing focus shifts b) uneven compression causing miscollimation 2, shear of pads/blobs a) by gravity causing mirror lateral shift causing miscollimation b) by uneven thermal expansion between the mirror and mount causing the mirror to flex and the wavefront to be distorted. 3, uneven support of the mirror causing sag in the reflecting surface and distortion in the wavefront (assuming more than three pads)
IMHO #2b is the main reason a thick blob of silly-cone is used, if any glue is used at all.
#3 is only a problem with large and/or thin mirrors requiring more than three support points on the rear of the mirror (ie. most of the interesting cases). It is deformations of type 3 which I think makes Phil Harrington's design very likely to fail, while his explanation seems to indicate that he has only taken steps to prevent deformations of type 1 and 2.
His cell design has a plywood sheet, thin high compression (does this mean hard or soft?) rubber gasket, and the mirror. Now, I do not mean to imply that Phil Harrington's mount is not working (right now). But suppose the plywood base, upon which the rubber gasket rests, warps over time and becomes one millimeter taller in the center of the mirror? What if the back surface of a different mirror is not flat to within a tenth of a millimeter. What if there is an irregularity in the gasket material which makes it one hundreth of a millimeter thicker in one spot? All these seem very likely. What will these do to the wavefront?
Assuming the rubber gasket is of uniform thickness, to have close to even pressure across the back of the mirror the amount of compression (the displacement or change in thickness of the gasket due to pressure) must be large in comparison with the size of the surface irregularities of the mirror and the front of the plywood cell. For small values of surface variation the ratio of the surface variation to the compression will be the roughly same as the pressure variation. Ie. if the surface irregularities are one tenth as large as the average compression, then the pressure will vary +/-10% over the back of the mirror. If there are surface irregularities greater than the amount of compression, then those points will support either all or none of weight of the mirror. Similarly, to allow for irregularities in the thickness of the pad, the amount of compression should be large compared to the thickness variation.
This indicates that a thick soft gasket material will distribute pressure more evenly than a thin hard gasket. I would think that you would want at least a millimeter or two of compression if everything seemed pretty flat and more if there were significant surface variations.
A thick soft pad will clearly make deformation 1a worse, and possibly make 1b and 2a worse depending on other mounting details.
So it seems to me that this design is not likely to result in a very good mount unless very flat surfaces are used and no warping occurs over time. Even then, the focus shift may be too much of a problem for some.
Well that is probably saying too much already so I had better stop before I get even farther behind.
Chuck