There's a subtle thing here that I think often gets lost in discussions of this nature. The fact is that much (most?) innovation is "top-down" in the sense that there is one person holding the entire idea in their head that ultimately drives its attainment. That person might be a team of one, in which case they are just managing themselves, or they might have 20 people reporting to them that they can direct.
Whether you consider the resulting innovation top-down or bottom-up really depends on the context of that person within their organization. (And if you are in the organization, it depends on your own position in relation to that person).
Consider a manager in a company like Google who has 20 people reporting to her. Imagine that this manager has a vision of some innovation she believes she can achieve through the work of her 20-strong team, and so she manages the team in an extremely hierarchical and directed way in order to achieve it. She sets goals for individuals, she approves all design decisions, she vetoes any aspect of the project - at any level - that she doesn't like or that don't fit into her vision of how the result should look.
If the result of this process is ultimately perceived to be some Great Innovation (say, something like Google Maps), then outside observers are very likely to point at this as an example of why "bottom-up" is the best way to get innovation. After all, the manager was low-level, and was operating outside the direct influence of upper management, such that the innovation "emerged" rather than was designed from the top down.
Yet this same scenario tweaked such that the manager is instead the CEO of a 20 person company suddenly looks like the epitome of "top-down" hierarchy a la Steve Jobs. People will point at the CEO and say that she is controlling and hierarchical. But, again, if the result is good, this will be used as an example for why top-down hierarchies are "good" for innovation.
I've witnessed this directly in my own career. Several years back, as the lead of a team of ~20 people, I developed "innovative" new products that were not dictated by upper management of my 2000-person employer. It was seen as 'bottom-up' innovation in the organization, even though I was fairly hierarchical with the team and driving them to my vision. No matter, it was 'bottom-up' because I was innovating without being instructed by my bosses. Flash forward to being CEO of a 40+ person company with a ~20 person product/engineering team. The same characteristics that brought me success and the perception of "bottom-up" success at the large company are now perceived as "top-down" and controlling in this organization.