The following is common for all Boeing installations - AB and Douglas/MD implementations are not necessarily the same.
The throttle snap back and power reduction is based on the actual T/R cowl position. If it's only 'unlocked', nothing will happen - the reverser needs to be ~10-15% deployed before the thrust cutback (and non-FADEC thrust lever retard) takes effect. On non-FADEC engines this is done by a cam in the strut box that is positioned by cables connected to reverser sleeves - when the reverser moves far enough towards deploy, that cam restricts the thrust lever position to idle. That same cam restricts the throttle to reverse idle during 'normal' T/R deployment - then allows full reverser thrust lever movement once the reverser is ~90% deployed.
FADEC uses electrons to do the same thing, the throttle doesn't move for uncommanded reverser movement in forward thrust. In reverse, the FADEC engines have a mechanical interlock that prevents reverse lever movement beyond reverse idle until the reverser is ~90% deployed (to give a tactile indication of T/R position). There is no throttle snatcher in forward, but in reverse if the T/R stows uncommanded, the reverser interlock will snatch the reverse lever back to reverse idle. Uncommanded stow is very rare, but if it happens watch your hand - it can easily trap it between the levers and do serious damage.
Post Lauda - with the addition of the third lock - the current cert basis for Boeing thrust reversers is that they won't deploy in-flight. No first hand knowledge here, but perhaps the AFM/QRH reference isn't there since it supposedly will never happen. Note that this is applicable to 'high bypass' engines - I don't believe the third lock was ever mandated for the low bypass engines (e.g. JT8D on the 727 and 737-1/200).