My three criteria for identifying life are: resists entropy or decay, and, does so without aid unless nearing death, and, secondly, life can process or respond to information about its surroundings, and the organism/species must be capable of evolving over time. Thus most extremeophiles (tardigrades or the like) fit the criteria. Extremeophiles such as the tardigrades (which are polyextremeophiles — they are adapted to multiple environments), still resist entropy and can respond to information (undergo cryptobiosis when the environment becomes too harsh). The other types of extremophile are as follows: Thermophiles (heat resistant), cryophiles (cold resistant), halophiles (salt resistant) and acidophiles. All these organisms fit those criteria mentioned above. However, the liquid droplets mentioned do not, as the droplets are not capable of Darwinian evolution. Another example of entities not considered life could be weather patterns, specifically storms or hurricanes. Storms can reproduce (form smaller storms by colliding with other air fronts) and can respond to the environment. Moreover, storms also resist entropy on their own so long as they have a supply of warm ocean water. However, once again, storms/tornadoes/hurricanes are not alive as these cannot undergo Darwinian evolution.

No, as you can still respond to the environment, if you observe something startling, your heart rate will increase. You will never be completely paralysed. There will always be some movement involved. If you go into complete paralysis and go into comatose, you can still breathe and thus are interacting and responding to the environment (that is what breathing is) so it still does not violate the definition I gave.