Description of problem:
The test seems to be failing specifically in 4.6 and 4.7:

4.7 jobs:

"Found in 34.30% of runs (49.17% of failures) across 172 total runs and 86 jobs (69.77% failed) "

https://search.ci.openshift.org/?search=an+end+user+can+use+OLM+can+subscribe+to+the+operator&maxAge=6h&context=1&type=bug%2Bjunit&name=4.7&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

4.6 jobs:

"Found in 9.38% of runs (12.00% of failures) across 64 total runs and 48 jobs (78.12% failed) "

https://search.ci.openshift.org/?search=an+end+user+can+use+OLM+can+subscribe+to+the+operator&maxAge=6h&context=1&type=bug%2Bjunit&name=4.6&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

it is not seen in 4.8 jobs:

https://search.ci.openshift.org/?search=an+end+user+can+use+OLM+can+subscribe+to+the+operator&maxAge=6h&context=1&type=bug%2Bjunit&name=4.8&excludeName=&maxMatches=5&maxBytes=20971520&groupBy=job

which makes me concerned it's not failing due to an external environment issue, unless the test uses a different operator for 4.7 vs 4.8.

sample run:
https://prow.ci.openshift.org/view/gs/origin-ci-test/logs/periodic-ci-openshift-release-master-nightly-4.7-e2e-aws/1412955830180909056

fail [github.com/openshift/origin/test/extended/operators/olm.go:261]: Timed out after 300.116s.
Expected
    <string>: 
not to equal
    <string>:


The debug/state info this test dumps out when it fails also seems like it could be improved.


Version-Release number of selected component (if applicable):
4.6, 4.7

How reproducible:
failing pretty regularly in CI