C6678 ‘Pacing Delay’ in SRIO throughput benchmarking example

I am trying to understand the use of the ‘pacing delay’ as found in the SRIO throughput benchmarking example installed by the MCSDK in

C:\ti\pdk_C6678_1_1_2_6\packages\ti\drv\srio\test\tput_benchmarking

I refer to the DIO NWRITE tests.

Referring to the doc below, I understand Pacing Delay  as ‘The delay cycles added to keep the RX side from getting overruns’

 C:\ti\pdk_C6678_1_1_2_6\packages\ti\drv\srio\test\tput_benchmarking\docs\SRIO_Benchmarking_Example_Code_Guide.doc

From the code, I can see that it is the delay intentionally added between the transfers of successive packets at the Producer.

Does the ‘pacing delay’ guarantee all transferred packets are correctly received at the Consumer? Is it not sufficient that the socket (using Srio_sockSend_DIO()) which is programmed to block until a LSU transfer is successful based on the ‘Code Completion bits’  guarantee a successful packet transfer as well?

Thanks.

Luke