Why Wan2.2 T2V workflow use HunyuanLatent node ?

[NielsGx]

The official workflow for Wan2.2 14B T2V use EmptyHunyuanLatentVideo and not Wan22ImageToVideoLatent (unlike the 5B workflow)

Any reason why ?
I tried to use the wan22 node above, plugged the VAE in it, but I get 512x512 video when asking for 1024x1024 🤔