FindTheRats, this one's for you.
TwitterRep <-- drag to bookmark bar
This code looks at a user's follower count, following count, and update count and alerts back a number. No saying how reliable it is (won't work for users with more than a million of any of those values, BTW), but it seems to do a good enough job of picking apart the wheat from the chaff.
mousavi1388 -- 712174233
iran09 -- 681975
TehranBereau -- 371922
persiankiwi -- 82745541
StopAhmadi -- 8964349
Change_for_Iran -- 149722592
simonjake -- 19
tahria -- 27
giti456789 -- 1
Toby1776 -- 18
iranianlecturer -- 174
serv_ -- 3898
Karmuk -- 89
As you can see, the users you should follow have TwitterReps several orders of magnitude larger than those you shouldn't follow.
And by the way, I've got a TwitterRep of 46. Stop laughing.
Note: Just because you have a low TwitterRep does NOT automatically mean that you are un-trustworthy, as my account is an example. Just follow th converse: you CAN trust the sources with TwitterReps over, say, 50000.
Sorry, this is wrong. Both TheEllenShow and CNNBRK have higher TwitterReps, CNNBRK winning with a TwitterRep of 638,462,456,808, over 6 TIMES aplusk.
Also for the curious: the calculation scheme I use is:
b = # of following
a = # of followers
u = # of updates
TwitterRep = log10(10+u)*a*(a+1)/(b+1);