True enough. Certainly a hassle for the people who used it.
Anecdotally, I was always surprised when I ran across it being used, as it was a relatively infrequent occurrence. It seemed to always work well though, when I did. I was even tempted to use it for a few projects. Guess I will have to seek other avenues. Be it s3, dropbox, cloudapp, droplr, or something else.
It still leaves some prominent open source projects, including ElasticSearch, with extra work that they need to do for their next release.