I believe exists because depending the sensorial modality in the enviroment, there is a need to find all the patterns, I think most of the time in laboratory tasks using just one or two controlated stimuli is hard to create the usual brain needs, for example you heard the keys when you are typing, see the letters, and the tongue move a little bit when someone is chatting, all that using different sensory information. So the delay depends of how many and what kind of information is around.
That's why I like this thesis: http://discovery.ucl.ac.uk/1310481/1/1310481.pdf
Now, I don't know the answer, I just think they kind of ignore or don't want to deal with the explanation of the sensorial difference and how the memory process is creating different paths for everyone...
I think you missed the point, hippocampal and semantic memory tend to be most measured at times greater than 1 minute or 60 seconds. I am talking about implicit memory not declarative, that stretches from about 3 seconds up to 1 minute. In other words after the implicit memory then comes declarative memory. The top limit probably has to do with short-term versus long-term declarative memory, and there is likely to be an overlay between them, but the implicit memory involved is still in the cerebral cortex where priming memory resides..
The issue is why ISN'T implicit delay memory being modeled.
"The issue is why ISN'T implicit delay memory being modeled."
Because like explicit STM, it is volatile, and difficult to map, and since it require practice to be converted into some sort of (skill), so, without repetition, it is lost and hence difficult to map. But given the very short time of priming, repetition of such implicit events utter difficult to interpret: but it can be mapped: Notice how guitarists develop such implicit skills, how they adapt with their notes by repeating such sequence. Violin and guitar players are best examples of such very short term implicit memory (VSTIM).