I wonder if they could've avoided this mess for API developers using UUIDs instead of integers for tweet ids.
Wow, means Twitter doubled in tweet volume in just a couple of months!
the interesting bit is how the quality of the average tweet could be measured and whether or not there is a strong relationship between volume and quality.
I guess it depends what you mean by 'quality'. A methodology for determining 'quality' would be awesome, some function of timeliness, authority, context, accuracy. I don't see that quality is correlated with volume though.