Knowledge of how to implement a search algorithm is pretty much useless in most real world applications in businesses especially when most people would just leverage what is already present in a framework like
.NET's linq or use the power of a RDMS to sift through data. There is often no need to "reinvent" the wheel and even if there was such a need, chances are, someone on the team would have already written a generic common library function for the most efficient search algorithm if you happen to be using a framework poor language such as C/C++.
If I may extend your use of "reinventing the wheel" a little further, it's like expecting a modern-day auto mechanic to know how to give a Model T a tune-up. Sure, he could figure it out eventually (especially if you illustrate it to him), but it's probably not something he's ever been exposed to let alone know off the top of his head. All of his normal diagnostic toolkit would be gone (computer, standardized gauges, timing gun, even the humble timing torque wrench).
iOS minimised the problem by limit[ing] the number of devices that developers need to target and test against
You were quite convincing until this statement. If we're down to splitting hairs, Apple does not "limit" the number of iOS devices. Apple clearly states that only Apple can make hardware to run it on and only Apple can make the tool chain to construct your application. If they allowed even just one other vendor, I would buy your phrasing. The way it is, that seems like an apologist way of saying "they enforce their own small-scale monopoly".
Physicist A: "We need to make something cool out of them... like one of those tiny violins or the art on a microchip..."
Physicist B: "Let's make boobs! Every sculptor makes boobs eventually! Quantum boobs!"
Physicist A: "OK, but they gotta be HUGE! Then we can use normal particles for the nipples!"
I've been hanging around lonely geeks too long.
Frankly, I like having the suggestions pop up (and not just for the fun factor). There have been times that a suggested result reveals the truth of something when the marketing and SEO have worked to whitewash the search results themselves. When people run into problems with a product, they will search for their problem rather than the marketing speak. I wish I could give my real examples, but I'm contractually/legally obligated not to. I'll contrive a working one instead (though the contrived one is not as solid as my real examples...).
Contrived example: Pop the words "MS Antivirus" into google search. "MS Antivirus" is a name of a piece of malware posing as security software. For me, the third suggested search is "MS Antivirus malware". Without having that there, the search results for "MS Antivirus" that declare it as malware are all below the fold. The results for "MS Antivirus malware" have the wikipedia entry for the malware itself as the first result.
Here in California, abolishing write-ins gets proposed every couple of years and there. Many states have some severe hoops to jump through before a candidate can be written in. Regardless, the funding in many campaigns for the two major parties ensures that the populace only really knows their names and not any information about "fringe" candidates. Even the people themselves cast allegations of "throwing away votes".
I ask: Do you know who you will write-in if your congress-critter votes to pass SOPA? Can you name who you will vote for instead to your critter when you complain/threaten?
After the Blue force was sunk, the game was ordered to begin again, with the Blue Team eventually declared the victor.
Interesting way to spin avoiding the issue of such an attack altogether.