Sorry about the "tuples", let me re-phrase it for your better understanding:
.rewrite file contains a list of rules. Each rule is composed by a) a regex expression and b) a replacement string.
This is a classical regex replacement mechanism, but using a simplified notation. The usual syntax is: /[regex expression]/[replacement string]/[options].
Rewriting is how the web works to provide friendly URLs. It's how Apache, nginx, and all others do it. Currently, if you want to have friendly URLs, you need to have a list of regular expressions that redirects to an index.html. Then, index.html can, for example, use the history interface to render the appropriate page.
This is the current standard. And in my opinion, it's better to have a mechanism that developers are already familiar with than to create a new one.