The Y-chromosome only really contains genes relevant to sex determination, with estimates of protein-coding genes ranging from 50-450. For instance the SRY gene, also known as the testis-determining factor , is only found on the Y-chromosome.
Phenotypes such as eye colour are coded on the autosomes (i.e. not the sex chromosomes), and as such are determined by alleles inherited from both parents. As Koustav Pal mentions in his comment, many such traits are multi-allelic (i.e. determined by multiple regions of the genome), and are therefore more complex than first appear; eye-colour can change over time, and there are many more complexities that do not fit into the simple "blue", "brown", "green/hazel" categories . This also applies to hair, so the child you mention could become more like their father as they grow older (or not!).
Because each individual posesses two alleles for each possible trait (one from their mother, one from their father - we each have two whole copies of the human genome in our cells*) a person can be a carrier for traits, such as blue eyes, whilst themselves having brown eyes. In fact the situation is so complex (>15 genes affect eye colour) that two blue-eyed parents could have a child with brown eyes  - the historical perspective was the brown eyes were dominant over blue eyes.
I have focussed on eye colour because it is a good case-in-point: there are clear genetic determinants and heritability, and it is more complex than the popular perception. Regarding hair (or skin) colour, single-gene mutations can cause pale skin and red hair , but beyond these cases the same rules seem to apply; that there are multiple genetic determinants (which means the Mendelian single-allele dominant/recessive model does not really apply) that affect the amount of eumelanin (which is dark brown) and pheomelanin (which is reddish) produced, with together combine to make the spectrum of colours we see in the population - it is not true to say that there are only blond OR brown haired people, as many people lie somewhere between .
This has not been an exhaustive review of these topics by any means, and there may be further evidence either way, but it seems quite clear that it is not unusual for offspring to not actually take the appearance of one or both parents. There are many traits you mention (such as build) that I don't go into detail, but there is emerging evidence that these are incredibly polygenic (caused by many small-effect genetic variants - e.g. height ) and as such could lead to the same phenomenon.
* well, most cells - red blood cells and gametes do not.