The 13.5kHz resonance is shifted downwards when using a longer insertion depth, which naturally happens when you combine the coupler with the metal ear canal extension or the silicone pinna and a typical IEM (ex 8kHz resonance). So above X frequency you won’t find a constant transfer fonction between the RA0045 and the RA0402. You can see an example of this below, using CSGLinux’s data (ER2SE being an example of a deeper insertion depth, IG955 a shallower one).
View attachment 287418
With a typical 8kHz insertion depth, the direct comparisons that I’ve seen so far tend to show only a moderate difference between the couplers up to 5-6kHz, of a magnitude that is insufficient to affect the observation that the use of a silicone pinna changes the measured FR quite a bit up to these frequencies, more so than the difference between the RA0045 and RA0402 couplers.