Shouty is usually an indication of poor tonal balance. The Pensils are kind of hard to screw up, but it you have over stuffed them to the point that the bass is suppressed, then you might sound shouty. I personally did not find the A7.3's shouty. A bit sibilant, but not shouty.
There are a little over 400 "break-in" hours on the drivers. That is, running them in open air at very low to low levels. They were just mounted in the Pencils for the first time today. Never pushed to "normal" listening levels before they were mounted.
I'm using 3 1/2" recycled denim insulation teased out to the width of the cabinet. Immediately behind the driver, I pushed out a cavity about 3 - 4" deep. Insulation goes about halfway down the cabinet, lower cabinet is clear of insulation. Insulation runs pretty clearly front-to-back except for the driver cavity. (I don't think the insulation has expanded to the back of the driver, but I don't know for sure.
Bass goes pretty deep and seems fairly well-balanced. I haven't taken any measurements.
I know Mark has made a huge point about gradual break-in, but I think that you may have taken that advise to an extreme. To break-in, the drivers do have to move. Turn them up to ~70dB at 6-8' and let them run a while. I don't think that your drivers are adequately broken in. They will sound harsh and shouty for a while.
That may be it, Bob. I can hear a difference this evening after playing them for the last 12 hours or so. I was checking the level and watching the cone movement. I'm getting a few millimeters visible movement at higher levels. Playing a Telarc now with some pretty good dynamics and bass (going to leave it on all night and into tomorrow). Will report back after some play time, but there are positive changes even today. (Might get a second pair of A7.3s since the allen wrench slipped installing one of the drivers. Just a ding, but it just looks ugly.)
I will be finishing the 10.2 Pencils Friday (today). Those drivers have about the same hours as the A7.3 so I'll run break-in on them some more over the next couple of weeks, a bit more gradually than the over the cliff workout the 7.3s are getting now.
From my recent experiments I feel that the teased out/thinned out thicker Ultratouch Denim insulation can make the sound a bit forward, vs something like Jute felt which has a more deadening effect. My experience is based on double layers of 6mm jute felt vs thinned out 2" recycled denim. Something to do with reflections coming back through the cone?
The denim felt that P10 uses probably will be different as that looks more densely packed.
Also as Bob said, maybe time to crank it up a bit.
I ran my MA 7s for about 150 hours in their cardboard box while I was building the Pensil cabinets - starting low, increasing to confortable loud using am radio to make sure there was no heavy bass. They sounded awaful when I finally put them in the cabinets, but I could hear them sweeten up over an hour or so, very small changes after that.
If I'm reading Bob correctly, I think he's suggesting volume and density of fill, particularly in the immediate vicinity of driver could be part of what's going on here.
IIRC, when I tuned my Pensils (a year and half ago) I ended up pulling some of the fill from the recommended amount, and used loose fiber dacron. Fiberglass, rockwool, dense denim insulation may be great for some applications, but Pensils might not be one of them.
And without riding one of my favorite hobby horses, the Pensils benefit from bracing, the absence of which can affect things in unexpected ways.
The drivers are starting to change. Now playing Telarc Firebird CD at medium level. Plenty to move the cone but looks like well less than 1/4" / 5mm total movement. Definitely not bottoming out. Definitely getting some serious exercise. Sound continues to evolve.