Reduce Crawler Load

I am using pathauto, so every node and term has an alias.

Because processInternalPaths in expire.api.inc adds every alias to the list of expired pages the queue in boost_crawler_expire_cache contains both the internal paths (node/nid, taxonomy/term/tid, ... ) as well as the aliases.

And so (almost) every page to be recreated by the crawler gets put two times into the queue which of course doubles the load of the crawler (I am not sure it this means that the pages are created two times or if the second call already gets the cached file).

Would it be useful to filter out the aliases in this function and how could we do this?

Or would it be better to get the Cache Expiration Module to not add the aliases (and would that be feasible)?