People with two working ears are able to "beamform" sound. This means that when sound arrives at their ears, the brain can use the phase difference between the signals of the two ears to do several remarkable things. First, and obvious, the sound can be localized. The binaural person is able to tell from what direction the sound comes from and how far away the source is. But, the binaural signal can be used for far more that that. "Broadband beamforming" means that all frequencies of sound are beamformed, so that the various frequencies contribute independent information about the signal. This process also allows the binaural listener to exclude "noise", here defined as anything other than the signal of interest. This means that in a crowded room, a binaural person can hear the person they are talking to, while simultaneously excluding the voices of the surrounding mob of talkers to a large extent. The binaural person can hear the conversation.
In contrast, a monaural person hears only one channel of the acoustic signals. No localization or beamforming is possible. All the sound signals come in as one jumbled signal, and no separation of those signals is possible. The difference between binaural and monaural is profound. The monaural person can barely hear the conversation in a crowded room - by using the larger signal of the nearby talker, or by unique frequency characteristics of voices. Localization of sound is not possible, so that when people see me at a distance and call out to me, I do not know where they are calling from - a 360 degree search for the signal must then be conducted (or I will just sit on the curb and wait for the caller to get closer.)
"Barely hear the conversation" means that only fragments of sentences or words are comprehended, and these must be sorted from noise or any other fragments of sentences or words from nearby conversations. The brain must work hard to comprehend the meaning of the sentence based on those fragments, and the context of the conversation.
Much of this discussion falls under the category of the topic of "psychoacoustics," a surprisingly challenging issue to understand; research is ongoing. For more technical information, see this article in Physics Today: "How we localize sound". Most of the discussion on psychoacoustics is irrelevant to those of us with unilateral hearing loss!
A room filled with people talking has the property that as more people talk the noise increases. And as the noise increases, people talk louder, which increases the noise, etc. Pretty soon people are talking quite loudly, yet still hearing about as well as they did when the meeting started and everyone was whispering to each other. This scenario is basically a nightmare to a monaural person...
Deafness on one side of course means that sound coming from that side will not be heard. This is actually rarely the case, because reflected sound will most often reach the good ear. With practice and intuition, a monaural person can often tell which direction the sound is coming from by using the natures of the sound and the environment. The notion that someone with SSD is in danger of being, e.g., hit by a car from his/her deaf side is rather overblown.
While in some circumstances, reflection of sound from environmental surfaces can help a monaural person, in some circumstances this can be a hindrance. A monaural person seeks to place himself in the right place and in the right orientation to best hear the sound of interest, while excluding the noise. Sometimes this involves aiming the good ear at a single reflective surface to enhance a signal, e.g., at a booth in a noisy restaurant, the good ear towards the wall is preferred. In a noisy, acoustically-reflective environment, the sound will seem to originate from all directions, defeating the strategy. This is the case with the noisy, echo-y cafe's that seem to be all too popular these days. Echo-y here means spaces that have great reverberation, i.e., reverberative environments present an added difficulty to the unilateral hearing person.