Ask HN: How important are arbitrary "technical" questions?

So I'm a young, recent Computer Science college graduate out of a fairly decent technical school. During my education, I gained an incredible amount of real world work experience through a Co-op program (alternating full-time work with a company related to major and academic semester).

I gathered a ton of skills working in a Java based web-stack which helped spawn interest in other web technologies and eventually support work towards numerous projects which I have either led, or been a "senior" member on. I spent a majority of my education discovering that, while more academic programming challenges are important, they are less prevalent in the real world and that being a creative, strong product developer was much more important. (Note: just because this is what I discovered doesn't necessarily mean it's right). I work in a start-up now and plan to work in the start-up environment for the foreseeable future.

I'm asking how important technical interview questions are because lately I've been on a bit of a job hunt and I keep running into this brick wall of arbitrary, rote, pointless academic problems being used a base line for whether or not a prospective employer considers me a competent software engineer/developer.

I've never made it a point to practice these problems because my work experience (from start-ups, to two man teams, to 20+ member small business) doesn't suggest that they are in any way important or significantly valuable to day-in day-out development and operations.

My question is directed to current start-up and small business employees involved in the hiring process: Are these questions really that important to you? Are these questions really the best way to judge the experience, proficiency, and talent of a new engineer trying to enter the industry? Have you tried/do you try other ways to evaluate candidates? If so, how much success (if any) have you had with them?

Simply put, I know the how. I'm quite familiar with that process. I'm less knowledgeable about the what (clearly, or this wouldn't be a problem), but my real concern is why?

Why does this hubris drive development teams to institute expectations and interpretations that aren't always true? Where's the data that proves that this process really works for small start-ups and/or businesses? Are we all just trying to emulate the Google/Microsoft/Palantir hiring processes? Or does it actually separate the best and brightest engineers from the cruft? Does it really help the business succeed, or does it just seem like it should because it appears to the de-facto standard?

It seems as though this is the only thing that is really holding me back from being able to enter a wide array of job opportunities. I've realized and understood that because so many people view it as a serious concern for hiring and development teams that I've started my own practice and "homework" intended to improve my ability to easily answer these questions moving forward.

Lastly, I'm more than receptive to criticism and/or suggestions about how to tackle this problem (e.g. great places to practice and "do the homework" for these interview tech. questions).