In the past, we have seen occasional confusion by webmasters regarding how crawl errors on redirecting pages were shown in Webmaster Tools. It's time to make this a bit clearer and easier to diagnose! While it used to be that we would report the error on the original - redirecting - URL, we'll now show the error on the final URL - the one that actually returns the error code.
In the example provided...
URL A redirects to URL B, which in turn returns an error. The type of redirect, and type of error is unimportant here.
In the past, we would have reported the error observed at the end under URL A. Now, we'll instead report it as URL B. This makes it much easier to diagnose the crawl errors as they're shown in Webmaster Tools. Using tools like cURL or your favorite online server header checker, you can now easily confirm that this error is actually taking place on URL B.
Server header checkers that I use provide information for the entire redirect chain, so this explanation doesn't completely describe the situation for me.
While the new reporting as described is more precise, arguably in some cases it might be more helpful to have both the source url and the url which returns the error listed... and simply better identified.