That’s just the legal limit, i think we’re looking more at microwatt range or lower. Somebody on stackexchange measured power output of phone nfc module, and - measuring power received at secondary of air-core transformer - got 0.8mW. Assuming it’s representative, if all that power was fed into phone-sized magloop (10 cm square made out of 6 mm copper pipe) then at this frequency (13.65 MHz) we’re getting some -30dB efficiency so 800 nW. This is already very charitable overestimate, because NFC antenna is smaller, thinner, printed and multi-turn, and does not behave like magloop at all. Any response from passive device would be much weaker still. All of that uses high data rate (wiki says that ISO/IEC 14443 says 106 kbps), and because these things are where people live, there will be also a lot of urban noise. I’m not saying it’s impossible, but it wouldn’t be that easy