I recently argued that asking for letters of reference to be submitted with applications for tenure-track positions devalues the time of highly productive scientists:
Letters of recommendation should be requested by the committee, and they should be requested after the first cut.
A senior researcher who’s a friend and colleague recently told me that for the last few years, she has simply not applied for any position that asked for letters of reference up front. It’s not hard to understand why. After a while, you start to feel that you’re imposing on your references (in reality, it’s the search committees who are imposing). In my case, I’ve submitted well over a hundred such applications, which means that each of my three references has supplied over a hundred letters (and thank you for that).