I believe that the prior art you are refering to is from Magar 1987 patent not 'Magnus'...  and I interpret this as you have.. the defendents using this citation in order to state that 'hey.. you just can't point to any  external crystal connected to plus and minus pins and say it infringes...  cuz Magar did this in '87...'

from MEI's Opposition to amend PICs:

7 Simply pointing to an on-chip clock generator, a connection of an external crystal to plus and minus pins, or a PLL, cannot be sufficient because those structures all existed in prior art, including prior art that TPL explicitly argued did not contain an entire on-chip ring oscillator. See, e.g., Ex. CC at 4 [‘336 File History, 7/7/97 Amendment] (distinguishing Magar prior art reference that disclosed an external crystal across plus and minus pins connected to on-chip clock generator).

Refering to:

In each case, TPL merely notes that a ring oscillator may be present in the chip, but does not actually point to a ring oscillator that is on the chip.7

I agree with you.. this is an attempt by MEI to state that obviously not every crystal on every chip infringes, there by attempting to limit their scope of infringement.. not necessarily screaming 'prior art!, their patent is invalid'... because, as you mentioned, this patent of Magar's was referenced in the original Shboom patent app. in '89.

Now, I believe, the unreferenced patent supposedly containing prior art was Ledzius'..  [who also was kind enough to pay us a visit here.. what a guy] this was cited by Pubepat in their reexam request to the USPTO.

corrections welcomed

regards