What bothers me most about memory is how misleading it can be.
Implanting false memories is in fact quite easy.
Matthias, so true.
(In this radio program, Elizabeth Loftus mentions how unpopular she and her research became.)
I think implanting false memories would fit with the illusion-of-truth type of implicit memory.
The program also goes into some detail about memory degradation, how in fact each time a memory is recalled, it has to be "recreated" in some ways. The participants talk about how the more a memory is repeated, the more it can potentially degrade. In effect the situation is that one doesn't really recall a "past" so much as one considers another reconstructed "now". Here is a link to an essay by Joseph LeDoux, explaining how his thinking on memory was changed by some work by Karim Nader that showed memory is reconstructed each time it is accessed.
This reminds me of Blakeslee's book, The Body Has a Mind Of It's Own. Even procedural memory (that of a skilled activity) can degrade. It may be more optimal to practice a movement (a golf stroke for example) mentally rather than physically, part of the time, to avoid an onset of dystonia.
So there we are. There is some sort of optimal amount to practice a behavior, to "learn" it well, to commit it to "memory": practicing it too much without enough rest will degrade it. Not knowing how to UN-do something and practice THAT as well, might be something to think about.