I want to go further about the point that Modeler_In_The_Mist brought up. In the screencapture, all of the monster's features blend together so it's hard to tell what's going on. However, this will be mitigated when it's moving, as it'll introduce depth and details via relative motion. Even so, confusing the silhouette is usually on a game artist's list of things to avoid.
There are two elements of silhouette that artists typically pay attention to. First is the physical silhouette, that is, the outline of the geometry of a creature from any view. If you render the alpha channel of the monster from multiple viewpoints, it should be obvious just by the black-and-white shape where the head, arms, etc are. Below is an example from one of my old game models I made.
(http://an8.the-meadow.net/penguin_silhouette.png)
As you can see, looking at the silhouette of the physical form in the top row, the features are somewhat obvious, except from some perspectives---if you didn't have the other views as a reference (this is partially what I mean by details via relative motion). What helps this is the texture, which has features on it that make it obvious what's going on. The hat is obviously a hat because of the white border and gray gradient. The arms are obviously arms because of the expressive yellow lines on them. The eyes are easy to pick out because they are outlined in white feathering, and the beak stands out because it's orange with a dark outline. The form of the back is aided by the expressive yellow lines that curve along it from neck to tail. All of these elements in the texture help its form.
However, your monster is confusing because the textures are very sporadic. The only thing really done right with it is that the claws are yellow and therefore easy to pick out. The real problem with the rest of the texture is that there are just random rings that contradict its physical form. Line patterns on objects are known to obscure the object itself, so why did you use it? And don't say that it helps the monster's design (creative license), because that's just a convenient (and wrong) excuse artists typically make.
The other problem with the texture is what's between the rings. Black. This just blends the depth of the monster. If you aren't using hardware shading in your style, nor using any expressive lines in the texture to aid the viewer's eye, then having just common solid colors without any sort of gradient shading drawn in only obscures the monster's form even more. Not to mention the black blends in with the ground.
The point I'm trying to make is that you have to put a lot of thought into the silhouettes, and especially more so with the textures if you're working with low-poly models. You can take this advice with a grain of salt, or learn from it. Whichever is entirely up to you.
I'd like to share a useful program, just in case you want to explore the next steps in texturing. XNormal (http://xnormal.net) is a standalone free program that allows you to bake ambient occlusion maps, normal maps, height maps, etc, using a high-res model to bake it onto a low-res, uv-mapped model. To bake AO maps on a low-res model without a high res one, all you have to do is set the low res model as both the high res and low res, then generate the map.
Anyways, good job so far. Last time I used Game Maker, there was no such thing as true 3D implementation :) As for the 27-second CG, typically in trailers the CG "wow" stuff is placed first to draw in the audience, and then the in-game footage driven in afterwards. But it kind of depends on the content of your CG clip as well.