What do you say to an almost perfect baby?

not so. With multiple and reduced genic expressions in phenotypic arrays, we barely are scratching the surface.

We have already affected the genes of future generations, in context of our epigenomes.
http://en.wikipedia.org/wiki/Epigenome

If I understand the concept correctly, the manner/environment in which we live changes gene expression in our inherited genome; however, these "expressions" can be passed on to future generations.

Theres another area ("Lamarkian" modes of inheritance" via epigenetic mutation or recombination) in which we have little idea how transfer is actually accomplished.

...I have have ear something about this neo Lamarkism before but very scarce information... do you have any source on the matter worth looking ?

We maybe barely scratching the surface of the inherent knowledge that is hidden in the world of genes, however, there is sufficient information for application matters such as altering a few human genes to give them desired properties. Similar to genetically modified foods.
For example China produces a new type of rice, named golden rice, that contains the genes to make iron. Such genetically engineered rice has the potential to prevent blindness in poorer communities where poor eyesight is common due to iron-deficiency .

Back to point, researchers in Cornell University added a green fluorescent protein to an embryo left over from assisted reproduction. This embroyo was destroyed a few days later.

As with genetically engineered food, new problems/ unknown problems await for us in the future.

So should science be taken much further to better ourselves (with a lot of trials and errors), although there are unknown risks in this endeavor?
Or should we leave it to mother nature's "natural" evolution to better our speciess at its own gradual time?

EV ERY short or long term wexpeiment with life has yielded unexpected results. As i chaos rules thegame. Weve made GM anything and created "Super weeds and super bugs". Weve cloned and gm modified organisms, none with expected results. I read genomics literature also.

You mean for cancer patients, with turning on of off a gene.

On the same subject but on a differnt note I also heard of the lazarus project where scientists are looking into extracting genes from DNA taken from the remains of specimens of extinct species, and perhaps bringing back recently extinct animals .
That is very interesting.

Put a nice suit on a Neanderthal, and possibly no one could see the difference from a modern human, except for the jutting forehead?

It is better to follow time tested methods of controlling genes; starting with mate selection. Check their grades, and give them a battery of intense psychological exams, called dates. Studies show inbreeding can lead to nasty genetic diseases that can be avoided if you advocate and pursue interracial marriage, and inter-tribal breeding. Otherwise like a German shepherd you can get bad hips or something like that. Already the Dutch Amish have rare genetic defects. Do not sleep with your siblings, even if you are a pagan from Hawaii, 2000 years is enough of that. Find the opportunity to check for health indicators, go dirty dancing, see if your potential mate has any attractive skills and talents before you sleep with them, so you can ditch them before they give you crabs.