Interesting coffeed discussion.
Even though I agree with bits and pieces of what everybody posted (in the thread), I think Tim Wendelboe pretty much summed up my opinions on this matter. I believe there is more than meets the eye when you examine a spent puck. A soggy puck does not necesary equate to bad extraction, and in fact, sometimes what appeared to be pinholes are actually not a result of channeling. You have to spent a little time playing with different combinations of setups before you can really see what the puck is telling you.
Here's my take:
I believe how wet or dry the spent puck is largely a factor of the head space (result of the combination of group head design, basket size/depth, and dose volume) and how the 3 way valve is plumbed.
When there is little or no head space, the puck's expansion during depressurization is restricted, which results in a nice firm puck. With the type of coffee, flavor profile, and shot pulling style of most US shops, you will see mostly firm spent pucks. In this case, pinholes and fractures will definately tell you that something is wrong with the dose/distribution/extraction (though side channeling and intra-puck channels are much harder to detect...).
On the other hand, if the total expansion is less than or equal to the head space, then the spent puck will most likely to be spongy and sometimes, soggy wet. This is due to the puck's free expansion. Now, how the puck looks like (especially the top surface layer) is another story. My theory is that, it depends on how the 3-way is setup, you will either get a nicely shaped spongy puck or an ugly roughed-up mess.
With the coffees I use (light-roast, high acidity/intensity beans), I found that I get a better flavor profile if I don't up-dose and keep a nice head space. On the Linea/GB5/Synesso, this results in shapely but spongy spent pucks (with the GB5 having the best looking of 3). On the other hand, on my Rituale, the spent puck mostly looks like a ugly mess. The top layer are often very disrupted with "hills" pulled up by the depressurization (and subsequently stuck onto and broken by the group screen). Sometimes, you even see craters which, w/out careful examination, could be interpreted as pinholes. The ugly puck issue troubled me for quite sometime. It continued to puzzle me as I have checked with a naked PF numerous times and there were absolutely no channeling or other extraction issues (and the shots tasted perfectly fine). I recently had Jaime double check me and was reaffirmed that the ugly pucks are not indications of dosing/distribution problems.
My conclusion is that, due to the setup of the 3-way in my Rituale, the depressurization at the end of extraction is much more violent than the LM/Synesso setup. As a result, instead of letting the puck expand (relatively) gently, the puck (especially the top layer) was ripped up which caused all the roughness of the top surface layer. However, this has no effect on the shot as the phenomenon occured after the end of the extraction. So just like what Wendelboe said:
"But really, If the taste is excellent, what does it matter ?"
- Ben
ps. food for thought (tossing a bone here, dear readers): I know it prob adds to the complexity of the mechanical design, but why don't we see a group head design that has a separate path for depressurization (or maybe it already exists and I am just not aware of it but will be interested in knowing...). This way, all the residue will be (mostly) contained in the depressurization path, leaving the brew path relatively clean. Sure, you still have to clean it to prevent any sort of clogging, but I imagine this setup will produce a better cup in a busy cafe environment (where you do not have to luxury to detergent-flush the group often).
