Despite the efforts of the GOP Establishment to proclaim Romney the winner and end the primary contests for the Republican nomination, Quin Hilyer wrote a detailed article in The American Spectator about why this race is not over yet.  Some excerpts:

No, no, no. The Republican presidential nomination really is not over despite the parade of rather craven endorsements now flowing Mitt Romney's way.
Before I go any further, a few words of explanation. Please bear with me. Part One of this column will explain why I am writing this right now. Part Two, the meat of it, will explain the actual delegate arithmetic in play, while also providing part of a tutorial on caucus/convention systems. Part Three will "game out" the remaining contests. Part Four will explain why a contested convention could be very much to the benefit of the Republican Party and its presidential candidate.


... It therefore galls me to see media accounts, and delegate counts, that misrepresent the possible or likely results of district and state confabs, which in truth can produce outcomes remarkably different from primaries. Because much of the "race is over" narrative relies on these potentially mistaken delegate counts -- and because about half of all the nation's Republican voters still haven't had a chance to weigh in on this contest, and will effectively be disenfranchised if the battle is concluded prematurely -- it is extremely important for both pundits and voters to understand where the narrative may be flawed.

The Santorum campaign has asserted that the delegate race is notably closer than the punditocracy has indicated. Their contentions largely have been dismissed by that same punditocracy, with some analysts describing the Santorum claims as inhabiting "fantasyland."
Here's why the truth is probably somewhere in between.

Part Two To understand why, one must understand that the differing assertions almost all involve caucus/convention states, and must understand that those states tend not to award delegates proportionally. Instead, they encourage results that approach "winner-take-all" decisiveness.
As a hypothetical, imagine a district caucus in which only two candidates are competing to choose, perhaps, five delegates to a state convention. Now imagine that 199 voters participate in the caucus. Imagine that exactly 100 of those 199 are firmly committed to one of the two candidates. If those 100 are disciplined, and can agree among themselves to let no more than five of their number run for the delegate spots, then they will win all five of those delegate spots regardless of what the opponents do. Why? Because all five of Candidate A's delegate designees would receive exactly 100 votes, while none of Candidate B's delegate designees could receive more than 99 votes. An almost perfectly evenly split caucus (100-99) would therefore produce a unanimous delegation to the state convention.
 
I highly recommend reading the entire piece.