I have included some basic information that may assist you in understanding the difference between a regular appeal and a de novo type situation.
Thank you mrkool, but I already understand the difference between a de novo appeal and a regular appeal as it is not a difficult concept. A regular appeal reviews errors made by the lower court whereas a de novo appeal reviews the arguments anew with no deference to the lower court's rulings (as if the previous trial never happened).
In the 1996 seminal case of Markman v. Westview Instruments, Inc., the US Supreme Court unanimously decided that claim construction was strictly a matter of law to be determined by a judge as opposed to a matter of fact, which is the purview of the jury (or the judge if a bench trial).
I'm not sure what point you're trying to make with the quote from former Federal Circuit Chief Judge Mayer. He argues against claim constructions being a matter of law and refers to the Supreme Court's Markman decision as "fiction."
The quote you posted without attribution is not authoritative. It is from his July 2004 written dissent on the court's decision to grant an en banc review of Phillips v. AWH which was a de novo Markman appeal.
Mayer, along with Judge Newman, argued that claim construction is a matter of fact much like the decision of obviousness. Therefore he did not agree with the de novo review granted in Phillips and thought the appeal should be reviewable for error only.
Despite his and Judge Newman's dissent, claim construction remains a matter of law rather than fact. In fact, Phillips became another seminal case in clarifying the hierarchy of evidentiary sources usable in claim construction: http://en.wikipedia.org/wiki/Phillips_v._AWH_Corp.
It is interesting that Judge Marcia Kreiger, who issued the Markman ruling in e.Digital's case, was the lower court judge who ruling overturned in the Phillips case, which one of e.Digital's lawyers tangentially mentioned during the hearing. In fact the clarity provided by the Federal Circuit in that case was significant in the e.Digital Markman ruling.