[feature request] node for region limited LoRAs / couples

Is there a comfy node for this? I searched the issues but couldn't see anything.

I've seen ways to compose LoRAs so you can have them apply to only part of the output render
This allows things like "couple shots" where you apply a different lora to left or right halves of an image, so maintaining separation on the two LoRAs. when i do this with a simple lora pipeline and two loras with different trigger phrases, there seems to be a lot of leakage between characters or the output loses it's likeness. This method of separation seems more promising for rendering "social scenes" with multiple subjects.

here are some relevant notes / plug-ins for the A1111 ecosystem:

https://github.com/hako-mikan/sd-webui-regional-prompter
https://github.com/opparco/stable-diffusion-webui-two-shot

https://stable-diffusion-art.com/regional-prompter/#Region_prompting_with_ControlNet
https://github.com/opparco/stable-diffusion-webui-composable-lora

older: latent couple
https://github.com/opparco/stable-diffusion-webui-two-shot

https://www.youtube.com/watch?v=vZ3W62dxuXI&ab_channel=fofraiartandexperiments
https://twitter.com/fofrAI/status/1640837861783007237