You can use both as per your choice, both works fine.
Implicit property node can be created by right clicking the control and create the property node and then selecting the property.
Explicit property can be created by taking the reference of the control and then wiring it to the propety node.
I recommend you to use the explicit property node as it saves the memory and reduce the file size if you are required to make the application for the real-time purpose.
Just for the simulation purpose you can use the implicit property nodes to reduce the complexity.
For a single control, I use Explicit simply because it uses less space and it is plainly obvious what it is connected to. If setting multiple controls, I like to use arrays of references and a FOR loop to autoindex over the references and set whatever properties. Obviously, this uses Implicit.
I am not aware about performance differences. Overuse of property nodes is a much bigger concern.
For altering single element there is no reason to create reference and connect it to the property node instead of a single node.
If programming or block diagram can be simplified by creating references and changing multiple elements at once - do it.
The other possibililty where you would want to use explicit property nodes is when you want to move the property nodes into a subVI, perhaps you have a lot of different properties to change, or you may need to do it for multiple objects or in different places in your code. Then you pass the reference or array of references into the subVI and let the subVI do all the heavy lifting. It will keep the higher level block diagram cleaner.