The rate of attacks in Iraq (for all types) has stabilized at the levels of 2004/2005, which were prior to the bombing of Askariya. During this earlier period, Iraq's open source insurgency was highly decentralized. A good way to quantify this is through the analysis of Oxford's Neil Johnson, who plotted the number of attacks against casualties per attack. He found that the equilibrium point for Iraq's conflict (as well as the long running war in Colombia) was a power law with a coefficient of 2.5 . A conflict with a coefficient of 2.5 looks more like intensive terrorism than conventional warfare. It is also likely, given Colombia's experience, a level sustainable over decades of conflict. See my brief "War's New Equilibrium" and Johnson's paper "From Old Wars to New Wars and Global Terrorism" for more on this.
The best explanation for the spike in violence between February 2006 and June 2007 is that the Askariya bombings initiated a process that was leading the conflict towards total war. Total war (Ludendorf) is a form of non-trinitarian conflict that ignores moral, political, cultural, etc rules in favor of complete mobilization to achieve total victory (global thermonuclear warfare is the ultimate example of total war). In Iraq, total war means religious cleansing via militias. Up until February 2006, Iraq was a limited conflict where US and Iraqi forces under strict modern rules of engagement, kept a lid on the scale of conflict (although they were unable to win). The Askariya bombing changed that dynamic. It so completely sundered the domestic social system in Iraq that the conflict lurched towards total war.
By early 2007, Sunni forces were suffering defeat after defeat as large Shiite militias violently cleansed towns and neighborhoods across Iraq (if measured in terms of Johnson's coefficient, we would have seen a move towards 1.8, the coefficient of conventional warfare). This put the open source insurgency in a crisis. Sunni insurgents weren't able to form the large local militias needed to defend themselves as long as they were in conflict with the US (these formations would be easy targets). The US military saw this opportunity and enabled the Sunnis for form local militias under the protection of the US (putting 60,000 on a US/Saudi payroll) as long as they sacrificed jihadi groups associated with al Qaeda. The US also began to target Shiite militias. The result was that the onrush to total war in Iraq was averted as the Sunnis began to develop conventional forces. The return to limited war also means that the open source insurgency can now thrive again.
An important rule of systems disruption that was ignored
With Askariya, the open source insurgency's attackers ignored Lawrence of Arabia's advice on the use systems disruption. Lawrence correctly argued that limited disruption (partial disruption) allowed the attacker to control the direction or flow of a system. Complete disruption forces fundamental changes to the system that will likely result in negative consequences for the attackers. Open source insurgents that ignore this will pay the price (as al Qaeda is today). This is from my June 2005 brief, "Partial vs. Complete Disruption":Complete collapse would create total war (via a bloody civil war). A complete urban/country takedown would prompt the state to launch a total war. This is a type of warfare that global guerrillas are not prepared or able to fight (in contrast, states are well suited to this). By keeping the level of damage below what would be considered fatal to the state, total war is avoided.