Dynamic loading and unloading of subpresentations during runtime

Allow for loading and unloading of subpresentations during runtime with a public API. This is to allow customers to create their own use-case optimised memory managers that keeps the RAM usage within customer defined parameters. The decision which subpresentations to keep in memory and which to unload in certain situations is highly use-case dependant and can even vary within customers product line from one variant to another.