I am glad you mentioned the noise from the ceramics striking the inside of the envelope. People do not really appreciate this potential issue, but I have seen instances of completed units being damaged in transport because the shell supports had enough flex to allow the ceramic to hammer on the inside of the glass envelope. I have even seen this happen with brand new electrodes right out of the box.....a nice tidy little crack right there at the ceramic. (I'm looking at you, SVP)
In the early days, the problem of a shell contacting an envelope and causing damage due to localized heating during operation was well understood and many early patents deal with the internal supporting structure. (Some examples in my collection have internal glass jackets and bead rings or other contrivances to deal with this.)
Years ago, I had a group of EGL electrodes that almost always failed during process, right at the ceramic. I documented my process, and sent a few failed examples back to them. They replied with a letter detailing a manufacturing defect they found in their production line (part of the centering pin, if I recall) and gave me two new boxes of electrodes with which I had no further trouble. (Ah, the days of real customer service!)
Today, I am using some non-ceramic EGL electrodes on smaller work (7 to 10mm) ....for 12mm and bigger stuff, I've been using the Neon Products short shell ones and some long shell Masonlites. In both cases I am finding them to be satisfactory. I do also have access to a variety of older electrodes and in spite of their shelf age, they seem to work fine as well and in some cases perhaps even a bit better than what we get now.