Like most things these days 'remote' or 'digital' towers are driven by the need for industry to sell its systems and the potential to reduce the costs of ATC services. It looks as if industry have been successful in persuading senior managers in ATC that this way of working makes sense.
On the basis that a controller can only perform one role at a time and still has to be trained for that role, I can't see much scope to reduce costs through staff cuts. However there is potentially a big saving if this way of working avoids the need to build a new tower.