It might have to do with the large number of fires associated with R44 accidents...
When I first attended the RHC safety course in the 1980s Frank mentioned to the class that he was very proud that the R22 had never had a post crash fire that was survivable (i.e. any post crash fire was associated with impact forces large enough that the occupant were killed by the impact forces). He mentioned at the time that he had copied the Hughes 500 egg shape and attributed that with great strength and crashworthiness.
I've noticed that the same does not seem to be true of the R44. Taking a quick look through the NTSB accidents I see all the following in which there was a post crash fire and either people survived, or as best as I could tell reading the report people might have been able to survive (it's not always clear from the report).
ERA10WA510, ERA10WA428, WPR10LA354, WPR09CA470, ANC08CA112, LAX08CA182, DFW08LA122, NYC08FA026, SEA07FA223, DFW06FA102
Remember a while back Robinson came out with the safety notice suggesting the use of Nomex flight suites? I think Robinson has realized for a while that the fuel tanks of the R44 are much more likely to spill than those in the R22.
While it sucks to have to pay to upgrade the fuel tanks (and I assume lose a little tank capacity), it appears to me that the R44 really does need some changes to address post crash fires - the change to flexible fuel lines and this Service Bulletin seem like good ideas to increase the chance that we can walk away from a survivable accident...