An ensuite is not a small main bathroom. It shares a wall, a ceiling cavity and often a door line with a room somebody sleeps in, and the problems worth spending money on are the ones that follow from that rather than from the layout.
Sound arrives by three routes and each needs a different answer. Airborne noise through a single-board-each-side stud wall carries conversation clearly; what reduces it is cavity insulation, added mass, and decoupling the two faces so they cannot drum against each other. Insulation alone is the cheapest of the three and the one worth doing while the wall is open. Then there is flanking, where sound travels around rather than through — and the door is almost always the leak. A hollow-core slab with a generous undercut defeats a well-built wall entirely, so a solid-core door with seals at the head and jambs is the single most effective change available. Third is structure-borne noise from the waste stack, which is a pipe problem rather than a wall problem: wrap or isolate the pipe, and keep it out of the wall behind the bed where the layout allows.
Ducting is the constraint an interior ensuite runs into. With no outside wall, the route crosses a ceiling or an attic, and the run is what decides the fan rather than the label on the box. Every bend and every metre of sagging flexible duct costs airflow. Rigid or semi-rigid duct, the shortest practical path, insulation where it crosses unheated space so it does not condense and drip through the bedroom ceiling, a slight fall towards the terminal, and a proper wall or roof outlet with a damper. Never a soffit.
Then there is the fact that the only way into the room is through a bedroom. Every sheet of board, every bag of debris and the tub itself crosses that floor, which in practice puts the bedroom out of use for the duration. It is better emptied at the start than worked around.