This is a dual post from another thread but is on point.
The IRC is quite specific under opening limitations for guards that only required guards need meet any of the sphere requirements, 4", 4-3/8" or the 6".
Yankee is correct that the 2009 IRC added a 36" directional edge measurement zone to the 30" height trigger point and thus has increased what I refer to as a drop zone. This took me a few code cycles to get adopted and the IBC & IRC committees both requeted I do the change through the CTC's work with guards.
Weather or not you agree with the 30" trigger point, really does not matter, the code specifies 30" and therefore unless changed during the adoption of the code by the AHJ, this is the trigger point.
Also note that the 2009 IRC has added the word required to the height section of the code. Hence NON-Required guards under the 2009 ICC published vrs of the IRC does not require a non-required guards height to be a minimum of 36".
However in any case no matter if larger openings or short height, the code is specific in not stating (REQUIRED) for the loads and thus all guards defined by the AHJ as being a guard need comply with the load portion of the code.
These changes are not IRC specific, but also in the IBC.
As to changing to the 4" sphere to stop heads being entraped was not a real reason, the reason for the change is the large amount of small children falling through the large baluster spacing, not getting stuck.
For those of you that have the interest and the time the CTC portion of the ICC website, if it is working and comes back to life, has a huge amount of documented information on how the change in opening limitations from 6" & larger to a base size of 4" with exceptions has reduced the amount of fall through accidents for very small and young children. Nowhere in that information, injury data or studies did I see a problem with heads getting stuck. Just another myth.
My spell check is turned off for those typo's I bid sorry.