Sure this is probably the only reason the bonds have been rated as junk. If Twitter had paid then there is little chance that the quants down in the basement could say what they really thought.

On the other hand, it's in the rating agency's interest to give poor unsolicited ratings, to "push" companies into paying for solicited ones.

This is posible, but I suspect it would work a lot better if the agency has a quiet discusssion with the CFO before doing the rating. Something like explaining to the CFO that you can't trust those quants to understand the sensitivities involved unless they have a framework to constrain their thinking - if they are left to their own devices they might even say what they think.

I'm the last person to defend ratings agencies, but even in a world where ratings agencies were paragons of neutrality and forecasting accuracy one would expect that to be the case simply because those expecting an investment grade rating due purely to a sound business model can lower their borrowing costs by ensuring their bond is rated accurately by at least one of the agencies.

Since the agencies don't rate everything, the best way to ensure that is to pay a fee, so if you're issuing debt without the rating you're sending a pretty strong signal your bond is junk, whether the agency chooses to bother with a "speculative" rating on just how bad it thinks your fundamentals are or not.