You should probably consider the utility of any kind of turnout before deciding which will offer the best utility. If you have more space than the original, or if you decide to modify the original, you might find, as an example, that you prefer the wider range of rolling stock that a #5 turnout offers over a #4.5 standard or a #4 snap-switch with the largely curved diverging route. If you ever decide you'd like to run a 2-10-2 over your switch puzzle, you'll need at the very minimum a #5, but for a brass version you'd want about a #6....maybe.
Some switches can be sprung so that they allow one route when advance through from the points end, but the points will move for the flanges when the train enters from the frog rails end. Those would be found in model layouts and on streetcar routes. Very sophisticated, and costly, software and electronics can detect trains using sensors, and with programming, can allow the train through a prescribed route by throwing the points using servo-motors.
A great many modelers like the idea of automating their layouts to the extent possible. It's a challenge, a puzzle, and it offers immense satisfaction if it all works okay. Others, including myself, prefer to 'line routes' using a bamboo skewer and flicking the points from side to side as part of the route planning process. It makes the movement of trains more involved and requires the operator to plan the moves in succession.