Really, the "mech" in the video is total crap, and the John Deere mech is a far more realistic mech prototype.

Many advancements in three key areas will have to be made in order for a practical mech to hit the battlefield.

The first area is power generation. The power requirements of a huge mech would be so great that unless some strange new "energy crystal" is discovered and some way of directly converting the energy of said crystal into electricty is developed, then only a compact and safe fusion reactor and again a method of converting the energy it develops directly into electricty would do the job. Either way, the technology isn't here yet.

The second area is locomotion. A practical mech would need an extremely powerful, durable, and fast-acting means of actuating huge legs supporting tens of tons to drive a mech across a battlefield at respectable speeds and with some measure of agility. I think the only thing we have today that with further development might do the job is the linear induction motor. However, there is a lot of advanced materials development going on, and some of it has to do with alloys that react to the application of an electrical current with physical motion. A material like that might one day result in some kind of immensely powerful artificial muscles which might do the job with less weight and more efficiency than linear induction motors, but that's still a long way off.

The third area is armor. As one might guess, weight would be a critical factor in the creation of a practical battlemech. Unless somebody invents an honest-to-goodness force field, major advances in materials that are both superstrong and relatively light that are suitable for use as armor will be necessary. Active defense systems which can partially offset the need for extra armor will also be important. There are the recent developments of the Trophy Active Protection System which destroys RPGs and missiles in close range to an armored vehicle by shooting them down, and the electric force field that can defeat RPGs specifically by zapping their copper penetrating jets with electricty. Further development and possibly integration of both developments into a comprehensive active defense system may reduce the need for certain types of reactive armor and thereby save some weight, but they won't mean squat against the kind of firepower that can be brought to bear by heavy armor.

Basically, I think that large battlemechs can and will be constructed at some point, but probably not for about 70 to 100 years at the least.

That's assuming that nobody makes some breakthroughs of only God knows how many orders of magnitude and creates a somatic combat vehicle like the Iron Man armor. Who needs battlemechs when you've got highly agile, extremely strong, heavily armed, and heavily armored flying infantry that can self-deploy anywhere from the depths of the ocean to the vacuum of space?