Great talk guys! It is strange that we have been chatting up grids in a topic and it looks like you are also really into grids at this time.
Please look at How HTM is supposed to deal with spatial invariance? and lend your insights into where I have this wrong.
Or right.
It continues here: How HTM is supposed to deal with spatial invariance?
and here: How HTM is supposed to deal with spatial invariance?
From a theoretical view on grids look at this: How HTM is supposed to deal with spatial invariance?