You raise an interesting question - 1) did researchers just not see the feathers until there was the idea that they might be seeing feathers? But I can think of a few other interpretations. 2) Researcher don't systematically review everything that they see in amber but instead systematically search for specific features and ignore others (which is only slightly different than #1), 3) feathers are rare enough that it required a searching an extensive collection, which previously wasn't available, and 4) there are very people in the field.

Quoting from http://www.wired.com/2011/09/dinosaur-feathers-amber/ :

> When study co-author Ryan McKellar first looked at the amber, the University of Alberta paleontologist wasn't even interested in feathers. He was interested in insects. It was only when he and Wolfe noticed odd, hair-like structures that they wondered if something other than bugs might be preserved.

I think that gives weight towards your #1, since it shows that (at least some) researchers will notice things other than what they are looking for, which is my #2.

Then again, this Atlantic article says:

> What originally got me interested in this project was a single feather fragment trapped in a spider web that had been given to Dr. Brian Chatterton. ... As part of this project I was able to go through the Royal Tyrrell Museum of Palaeontology collection of 3,000 or so inclusions in great detail. After finding a lot of feather fragments in this collection, I had the good fortune of meeting the Leuck family, amateur collectors in southern Alberta, who donated a pair of additional specimens to the University of Alberta.

Wikipedia reports that "Only 11 specimens are currently known", so 8 feathers were found in a set of 3,000 pieces of amber with inclusions. I wonder when Chatterton identified that feather. And when the Leuck family recognized that they might have feathers.

These factors give more weight towards #3 and #4.

I can't figure the balance between the different factors, and the primary paper (at http://www.sciencemag.org/content/333/6049/1619.abstract#aff... ) is behind a paywall. But I am leaning towards your hypothesis.

There's a bunch of stuff in university and museum collections that come to light when an interested researcher goes to look for other stuff.

Recently some Charles Dickins notes were found that have been used to identify previously unidentified authors. Researchers didn't use any complex software to identify previously anonymous writing; they just read the notes that had been hidden in a collection.

http://www.independent.co.uk/news/uk/home-news/charles-dicke...

Some of the oldest fragments of the quran were recently found in a UK university collection.

http://www.aljazeera.com/news/2015/07/world-oldest-quran-man...

There's so much stuff in collections waiting to be discovered.

I imagine museum amber was not originally collected scientifically. The most beautiful examples were preserved - those with large insects in them. Any cloudy or obscured specimens got discarded as unworthy. That would be, any with unrecognized feathers, dismissed as flaws.