The Texas primaries are not a great comparison because of all the redistricting disputes in both the 90s and the 00s. Six of those eight primary elections you list involved either entirely new maps or residual disputes because of recent map changes. Moreover, several districts in Texas are divided between two roughly equal 'power centers' with a tendency to result in primary clashes between them.
None of that applies to NC-11, which has been geographically stable for decades. The political cultures are also very, very different between NC and TX.
This from Charlie Cook in an email, which suggests maybe why the RNC chairman has pushed the panic button. Holding the midterm election right now, might be highly unfortunate for the GOP (the numbers are the numbers):
"In the other variation of what has come to be known as the generic congressional ballot test, when people were asked whether they planned on voting for the Democratic candidate for Congress or the Republican, Democrats led by 12 points among adults, 44-32 percent; by 13 points among registered voters, 45-32 percent; and by a whopping 18 points among those most likely to vote, 50-32 percent.
I'm sure that's true. However, it wasn't my intention to apply it to NC-11, but to Armor and his comment about how hard it is for any candidate to take a 20% bite out of the hide of any [long-term] incumbent.
He's obviously wrong, based on years of data from all over the country. Texas was just an handy example. Am I going to spend the hours necessary to catalog the rest of the country and make a detailed statistical argument? No, he ain't worth the trouble. The raw data is out there to see for anyone who cares to look.