Does making graphics into unions increase performance?

The project I am working on has many graphics taken from the symbol factory. I am curious if making a graphic into a union would increase performance? A motor graphic, for example, has many shapes and nuances that make it a motor, but then turns into one shape when unionized. I would assume that since unions are one singular graphic, their effect on performance would be different than a graphic from the symbol factory.

Another small question I want to throw in. Does using gradient colors on graphics affect performance at all vs solid colors?

Union-ing a shape would probably have a fairly significant effect.
A gradient will, technically, require more ‘effort’ to paint than a solid color, but not very significantly so.

Thank you, appreciate it.