Thank you, Edwin for your interest in my essay,
1. Actually I postulate possible nullifying of observable effects not of the gravity only, but of all objective changes, i.e. all "natural forces", etc. It is a pretty bold (crazy enough?) statement, of course. The very possibility of that is based on this vague concept of "subject" , that can actually be any part of the universe, even be identical to the universe at the extreme. In such an absolute case it is easy to see that if "you" change the subject to reverse the changes ("back in time" ) it becomes possible. More practical particular cases are much more difficult to imagine, of course.
In case of gravity I was appealing more to Einstein's example of "reversing", i.e. nullifying of the local acceleration (local "gravity force"). Tidal gravity forces, of course could not be nullified this way, because they are of the "second order of locality", pardon my French. Though more deep thoughts on that particular case are definitely needed, may be this is exactly where back in time changes have to be applied.
2. This one is a kind of question that goes very, very deep. You see, the universal change caused by time (self-interaction) is not just some abstract change. First of all it is a change that we are used to term as "ordering" of events; next come some other, more particular changes. The whole process resembles the process of living cell differentiation, if you know what I mean. New physical interactions ("forces") appear with the development, differentiation of the "cell"-universe. There is more on that in my Philica.com article (see the link in References, check the Cached link).
3. No, I meant the very real time travel, influence on events through time. Any such experiment will cause measurable changes not only to the clock that we send back in time, but to the whole world that will be obviously changed by this experiment. Such changes, according to all that was said in the essay, would precisely fit the "energy" interpretation, so familiar to every one of us. I mean that such an experiment will release a certain amount of energy (but may be less that we wasted on it to conduct).