I'm not sure if that would apply directly to your proposed experiment, but my concern is that what we experience as memories can sometimes apparently be generated on the fly rather than accessing something that was there before.
In general, these kind of phenomenological studies suffer from subjectivity, and are very difficult to connect to explanatory mechanisms, without additional objective evidence. One relevant approach to deal with this is Dennett's Heterophenomenology[3], which seeks to combine the subject's own impressions with other external evidence - so in your case, instead of having the sitter do nothing, you may want to pay them to gather additional objective evidence.
[0] https://en.wikipedia.org/wiki/Confabulation
[1] https://en.wikipedia.org/wiki/False_memory