It’s not like the IRS needs a way to flag the new groups that were created in the wake of the Citizens United decision. They have all the information they need to do that without any special filter. They can search for the date of the application. If what you’re concerned about is that most of the new groups being created are in fact thinly disguised electioneering vehicles, then what you want to do is take a random sample of the new groups, review them, and see what percentage turn out to be self-dealing or otherwised engaged in inappropriate behavior.
Instead, the IRS method for dealing with the volume was to take an unrandom sample. And how did they decide that you deserved extra scrutiny? Because you had “tea party” or “patriot” in your name. Since the Tea Party was a brand new movement in 2010, they couldn’t possibly have had any data indicating that such groups were more likely to be doing something improper. So how exactly did they come up with this filter? There is no answer that does not ultimately resolve to “political bias”.
If Tea Party groups really were driving much of the post-Citizens-United explosion, there was no need to specifically search for the words “tea party” or “patriot”, because those words would naturally be overrepresented in a random sample of new applications. The reason you specifically search for those words is that you want to target those groups specifically, and not, say, applications with “Progress”, “Organizing”, or “Action” in them.