Pre-generating multiple text sprites vs. generating sprites as needed

Platform: Visual Studio - LUA
OS: Windows 11

So in this clip I put in the discord show and tell, we are generating four sprites, pulling the information from reference files and controlled by the position selector at the top of the screen.

I've noticed that on hardware, there is a little delay when swapping back to position one. Understandably, this is because it's the only position and card with some fake data on it for the demo.

The end project will probably have less text per position and card than the stand-in Lorem Ipsum we are using, so the delay might not even be present or at least not detectable in the end.

But... It got me thinking...

Instead of using these 4 dynamic sprites, being destroyed and created when navigating between positions, could we generate all the sprites, when entering reading review screen, required for the full tarot spread and just display them as needed?

I could have a loading screen of sorts display for a few seconds to give the system time to create all the sprites required and then we just swap the displayed sprite when changing the position selector.

So we would be front loading the work instead of as needed. Max sprites, for a Celtic Cross spread like in the demo, would be 18 text sprites.

Does this sound like a valid, rational, idea or are there other factors that makes this a terrible idea? Being my first project, there's plenty of limits and nuance that's unknown and lost to me.

Thank you for your time and consideration,