Here are the parameters:
I looked at all WRs/TEs that scored at least 300 FPs from 2003-2004, combined. PPR scoring, no rushing stats considered. All WRs/TEs must have played at least 24 games, and been on the same team in 2003, 2004 and 2005. I then wanted to see which WRs did really poorly against a divisional opponent in those two years, and see how they did in 2005.
The lowest a receiver scored against a division opponent was 48% of his average FP/G. Hines Ward averaged 15.1 FP/G from 2003-2004, but only 7.3 FP/G against the Cleveland Browns. In 2005, Ward scored 49.9 FPs in two games against the Browns.
Rod Smith averaged 13.3 FP/G from '03-'04; against the Oakland Raiders, Smith averaged just 6.1 FP/G. Smith scored 30.1 FPs in two games in 2005 against the Raiders.
Eddie Kennison averaged 13.0 FP/G; against San Diego, he averaged just 6.8 FP/G. He scored 28.4 FPs in two games against the Chargers in 2005.
Those are just the first three receivers I saw. The table is way too ugly to paste here, but I'll answer any questions you like about specific receivers.
Conclusions
Splits happen. In four games out of thirty-two, picked at random, there would be a quartet of really ugly games for a star. I'm sure some stud RB has stunk it up in weeks 5 and 11 of the last two years. There's no logic behind that, of course.
...
Here are the next three on the list.
Donald Driver: 12.4 FP/G, 6.5 FP/G vs. Chicago; 31.1 FPs in two games in 2005.
Marvin Harrison: 18.3 FP.G, 9.8 FP/G vs. Houston; 43.3 FPs in two games in 2005.
Jimmy Smith: 13.8 FP/G, 7.6 FP/G vs. Tennessee; 22.5 FPs in two games in 2005.
That's six examples now, and the six most extreme: all six receivers averaged fewer than 60% of their season average FP total against these division rivals. Five of them dominated that same opponent the next year, and one performed about average. Gonzalez was at 56%; history seems to indicate that he's more likely, rather than less likely to do well against the Broncos.
...
FP/G = FP/G from 2003-2004
Riv = FP/G from 2003-2004 vs. specific rival
OPP = specific opponent
NFP/G = FP/G from 2005
NRiv = FP/G from 2005 vs. specific rival
Name FP/G Riv OPP NFP/G NRivHines Ward 15.1 7.3 Cle 15.5 25.0Rod Smith 13.3 6.5 Oak 14.5 15.1Eddie Kennison 13.0 6.8 Sdg 13.0 14.2Donald Driver 12.4 6.5 Chi 14.9 15.6Marvin Harrison 18.3 9.8 Hou 17.8 21.7Jimmy Smith 13.8 7.6 Ten 13.0 11.3The same arguments against starting Gonzalez could be made against starting Donald Driver against the Bears, or Marvin Harrison against the Texans.
You gotta go with Lynn Swan here. 
But seriously, for me, the sample size just isn't there. Four games isn't going to tell me anything, especially when one of them was good.
I have no doubt you owned this guy in every league you played in during the late 80s early 90s...
