The domain "all my dreams and all TV shows" is OK. Concerning the other question, yes, you should universally quantify $x$ in $O_x \rightarrow B_x$. Otherwise you are left with a free variable and the formula is not a sentence. That is, its truth value depends on the value of $x$.
Finally, if the conclusion follows from the premises, the conjunction of the premises and the negation of the conclusion is unsatisfiable.
However, if you have two distinct elements in your domain, one is a black-and-white dream of yours and the other is an old black-and-white TV show, then you can satisfy the premises (all dreams are B&W and so are all TW shows) while also satisfying the negation of the conclusion. (No element of the domain is both a dream and a TV show.)
Note that formally you can use $x$ in both premises, but you are better off using a different variable in the conclusion: $\exists y (D_y \wedge O_y)$. It's easier to avoid confusion, even though in the end you'll be checking for the satisfiability of
$$ \forall x ((D_x \rightarrow B_x) \wedge (O_x \rightarrow B_x) \wedge (\neg D_x \vee \neg O_x)) \enspace. $$
You may wonder why is it OK to have just one dream and one TV show in the domain. The explanation is this: if there is one structure in which the premises are true, but the conclusion is false, then the conclusion does not logically follow from the premises.
In another structure (you may say, in an alternate world) both premises and conclusion may hold, but the existence of a structure like the one above says that there is no logical consequence.
