Use *iB for power-of-2 bytes units consistently

[MichaelChirico]

Split off from #7074. I believe @aitap did a related fix recently (not easy to search for it). But we should try and be consistent about: 1024 bytes: kibibyte=KiB, not KB (which is 1000 bytes), and so on.

Examples found quickly:

data.table/src/fread.c

Line 2797 in 0afb20f

DTPRINT(_("%8.3fs (%3.0f%%) Memory map %.3fGB file\n"), tMap-t0, 100.0*(tMap-t0)/tTot, 1.0*fileSize/(1024*1024*1024));

data.table/src/fwrite.c

Line 794 in 0afb20f

STOP(_("Unable to allocate %zu MB * %d thread buffers; '%d: %s'. Please read ?fwrite for nThread, buffMB and verbose options."), // # nocov

More to search through (lots of false positives since GH search is not case-sensitive, but a local grep will do better)

https://github.com/search?q=repo%3ARdatatable%2Fdata.table+%2F%5Bkmg%5Db+%2F+-path%3A%2F%5C.po%2F&type=code

[MichaelChirico]

Two particularly problematic things:

The use of buffMB as a parameter in fwrite():

data.table/man/fwrite.Rd

Line 17 in 4773ef9

buffMB = 8L, nThread = getDTthreads(verbose),

The use of mb as a parameter and output column name MB in tables():

data.table/man/tables.Rd

Line 12 in 4773ef9

\item{mb}{ a function which accepts a \code{data.table} and returns its size in bytes. By default, \code{type_size} (same as \code{TRUE}) provides a fast lower bound by excluding the size of character strings in R's global cache (which may be shared) and excluding the size of list column items (which also may be shared). A column \code{"MB"} is included in the output unless \code{FALSE} or \code{NULL}. }

I'm not sure it's worth a breaking change for renaming any of those, but something to consider. We can at least correctly describe whether they actually refer to MB or MiB (IINM it's always actually MiB).