It's the same amount of pixels though, just with reduced bitrate for unfocused regions so you save time in encoding, transmitting, and decoding, essentially reducing latency.

For foveated rendering, the amount of rendered pixels are actually reduced.

That depends on the specifics of the encode/decode pipeline for the streamed frames. Could be the blurry part actually is lower res and lower bitrate until it's decoded, then upscaled and put together with the high res part. I'm not saying they do that, but it's an option.

It’s the same number of pixels rendered but it lets you reduce the amount of data sent , thereby allowing you to send more pixels than you would have been able to otherwise