Sums of 4 (or generally $k$) unit fractions are by no means trivial.
There is a general criterion to Y. Rav (On the representation of rational numbers as a fixed sum of unit fractions. J. Reine Angew. Math. 222 (1966): 207-213.)
http://gdz.sub.uni-goettingen.de/dms/load/img/?PPN=GDZPPN00218186X&IDDOC=253620
The equation $\frac{m}{n}= \frac{1}{x_1} + \cdots + \frac{1}{x_k}$
is certainly soluble for $m \leq k$, but for $m>k$ the
same type of problems arise that one has for
the Erdos-Straus equation.
One can expect that for fixed $m$ and fixed $3 \leq kN_{m,k}$ k < m$, there is some finite bound $N_{m,k}$ such that $n>N_{m,k}$ admits a solution. But this is an open problem.
One can prove that for "almost all" $n \leq N$. The strongest version of "almost all", and a discussion of the parametrization of such equations is in C. Elsholtz, Sums of $k$ unit fractions, Trans. Amer. Math. Soc. 353 (2001), 3209-3227
http://www.ams.org/journals/tran/2001-353-08/S0002-9947-01-02782-9/
