XSS Prevention From Unsanitized Server Input

I have been working recently with DataTables and server sided processing. Some of the test data I was using was not sanitized ie:

<script>alert('xss');</script>

and discovered DataTables directly injects data into the innerHTML of an element. I did not want to waste time processing the data on the client side so I simply reworked the creation of cells. For example

                // Need to create the HTML if new, or if a rendering function is defined

                if ( !nTrIn || oCol.mRender || oCol.mData !== i )
                {

                    //nTd.innerHTML = _fnGetCellData( oSettings, iRow, i, 'display' );

                    nTd.appendChild(document.createTextNode(_fnGetCellData(oSettings, iRow, i, 'display')));

                }

Or this example in sorting:

            if ( /*was just sTitle*//*document.createTextNode(column.sTitle).innerHTML*/column.sTitle != cell.text() ) {

                //cell.html( column.sTitle );
                while (cell[0].lastChild) {

                    cell[0].removeChild(cell[0].lastChild);
                }

                cell[0].appendChild(document.createTextNode(column.sTitle));
            }

Note the replacement is just appending a text node. Some additional code was required to ensure the nodes are empty but it seems to be a relatively simple (and beneficial?) fix.

How can we go about implementing these changes? Is it something that would be possible to do for a nightly build that might work as an initialization option?