Example 1: Anecdote (couldn't find reference online) Designers of astronaut suits struggled with knee/elbow joints... until someone ELSE told them of the armor of (I think) Henry VIII.
Example 2: An irrigation system fails because the designer had not known about phenomena (and thus their countermeasures, already in use elsewhere) that occur in those specific circumstances (geological, or wildlife, etc).
Example 3: [edit: removed, misleading]
Example 4: Searching for inexpensive but hard containers for microsculpture tools / works-in-progress, spending months rejecting tool boxes and solutions for modellers and hobbyist jewelry makers, finally settling for a DIY solution... and years later learning of the very low wholesale price of single-use medical equipment like PP vials, especially Eppendorfs
Example 5: Product designers trying to avoid infringing on patents: must first find anything potentially related before addressing it.
5.a) Being sued because the patent infringed on was on an algorithm used in a totally different branch of industry
Example 6: The intention behind an "Is there anything else I should know about" question
Example 7: Trying to find out whether "an" is the correct artcle to use before the quoted part in Example 6, without knowing the name of the grammatical role that part plays, whether it is an exception, etc - reducing the effort to browsing in hopes of finding a similar enough example
Example 8: Spending 60 hours searching for the official name of this phenomenon, asking around, reading Wikipedia (starting from "logic", "information", "knowledge", "paradox", ...), judging it salient enough to be named - though just maybe not, and how am I to know?
[edit] The emphasis is NOT on unpredictability, but on information/knowledge already extant, known to someone, somewhere, being very difficult (short of chance or external input) to find because you can't know the questions to ask without already having some knowledge of the answer.