This question seems to me to need 4 categories of answer:
1) Mathematical
2) Linguistic/philological
3) Real world
4) Philosophical
As such the odd numbered categories are somewhat in antagonism with the even numbered ones, as should become clear below.
Note that this is a two-part question, and I have generally reversed the order of sub-questions.
This is largely because the second part (is this always true?) is significantly more complex and interesting than the first,
and that the first part (why?) is pretty much satisfactorily answered purely in category 1) (mathematical) below.
Anyway...
1) Yes, by definition (assuming your question was solely about positive quantities, as is very strongly implied by its wording)
2) Not necessarily, due to possible variation in interpretation of the language of the question. For example, there could
exist a culture in which "5" is the symbol for our "4", and vice versa; or where "true" actually means "false".
3) _Almost_ always, because of the large math effect in real life. I.e. one is almost always comparing like with like, in order
to produce a meaningful numerical comparison. However, there are exceptions, e.g. 5 US dollars < 4 UK pounds
4) Almost _always_ , because of the large math correlation in philosophy and its logical basis. Thus, an axiom in math <=>
an axiom in logic <=> an axiom in philosophy, to a very high degree of certainty. Mathematics stands as the very highest,
untouchable pillar of science and philosophy, and "5 > 4" is a "universal affirmative" proposition in logic. This means, that,
in the absence of specifically unfavourable interpretative effects, it will _always_ be true.
However, this begs an investigation into the nature of "truth" itself.
Logic, and logical reasoning, is the most fundamental branch of philosophy. But in as much as philosophy ever goes outside of
logic, there can be exceptions. Sanity=logic=good reason, but to a baby or a madman ignorance could be paramount (though this example
blurs the boundaries with the philological distinction).
Descartes was among the first to describe (if not recognise) this "scepticism", in contrast to Plato, who had dismissed sense
info in favour of reason ("dialectic"), and Socrates, who had placed arithmetic and geometry very near (though not quite at)
the very foundation of logic and philosophy (e.g. in Theatetus).
The "ultimate" truth may be a synthesis of the Platonic and Cartesian approaches, as developed by the antagonistic stances of
Locke vs. Berkeley, and resolved to a degree in the work of later philosophers like Hume.
For further info I suggest you consult a philosophy text on the theory of knowledge or "epistemology".