fix a bug iter_subproc

This commit fixes a small bug in `iter_subproc.__init__`.
The parameter `chunksize` was never used. Instead a
constant value was used. The new code passes the value
down to `iterable_subprocess`.

datalad_next/runners/iter_subproc.py

@@ -53,5 +53,5 @@ def iter_subproc(
     return iterable_subprocess(
         args,
         tuple() if input is None else input,
-        chunk_size=COPY_BUFSIZE,
+        chunk_size=chunk_size,
     )