There was a time when publishers would look at unsolicited manuscripts "submitted over the transom." ("Throwing" the manuscript sounds unnecessarily violent.) Such manuscripts went into the slush pile, which is to say, the pile of unsolicited manuscripts, not manuscripts that were automatically rejected. And sooner or later someone did read all
those manuscripts -- although only the first few pages if the manuscripts were obviously
bad -- and everything that looked promising eventually came into the hands of an editor. Some books from the slush pile did
get published. My first book is one example.
Now there are so many manuscripts, it's impossible for publishers to look at them all, they are filtered through agents, and those agents have to make an initial assessment through query letters or small samples (depending on the preference of the agent). This can take a long time, because all of these people have obligations and commitments to the writers they have already decided to publish or represent. But they have
to discover some new writers, because old ones die or stop writing, or no longer write books that will sell. So each publishing house does
publish a few debut writers a year.
Do writers with publishable manuscripts sometimes get lost in the shuffle? Of course. There is no way around it. There is no use being bitter about it. No one sets out to persecute new writers. They have too many pressing demands to waste their time on sadistic games.
Much less that "agent" being successful with landing a proper publishing house.
I'm not sure what those quotation marks around agent mean, but we'll leave that aside. Once an agent agrees to represent a writer, he or she works zealously to land a contract. They don't get paid unless they do land that contract, so it would be absurd to suppose that they don't try very hard to make that sale. Since they do take on so many new writers, of course they can't sell every manuscript. They take that risk with every new writer they sign up: that they will work hard to sell that manuscript without, in the end, receiving a penny in return. Of course they have to be selective -- which means using their best judgement, not that they will be infallible. .