Summary is most of the issue. I stopped by my site to find everything 500'ing. Looking in the log, the issue is rooted in the (ab)use of a TestRequest in blog to generate the URL for a post that is about to happen. (This is in response to a trigger.)

Here's a trace:

vendor/rails/activesupport/lib/active_support/dependencies.rb:100:in @const_missing'
app/models/blog.rb:183:in @request'
app/models/blog.rb:156:in @url_for'
app/models/blog.rb:168:in @article_url'
app/models/article.rb:26:in @location'
app/models/article.rb:40:in @really_send_pings'
app/models/content_state/just_published.rb:42:in @send_pings'
app/models/article.rb:67:in @send_pings'
app/models/web_notifier.rb:5:in @after_save'
app/models/content.rb:189:in @publish!'
app/models/trigger.rb:28:in @before_destroy'
app/models/trigger.rb:12:in @fire'
app/controllers/application.rb:15:in @fire_triggers'

The resolution in my case was to delete the triggers from the table and not send the pings.