I looked at your video, I see what the others were saying. The problem is that it's easy to see the flaws in your video (simple textures, visible tiling, simple shaders, lack of water, lack of vegetation) but not the advantages (real-time huge terrains).

My suggestions are:

1. Show off the hugeness of the terrains, both from a FPS and a rapid flyover view.
2. Show something built using your terrains, which is a bit prettier and more interactive than your editor view. This is what the user will aim to create using your tool. Include some vegetation, some animals, a built-up area that you cleared a space for using the editor.
3. Show the versatility of your tool. Think of some real-world places that it can achieve similar effects to, e.g. Mars, Tuscany, the Himalayas, the Sahara, Antartica. Do a tutorial or two of these on your site.
4. Maybe pretty up your in-editor view. Fix the tiling by using larger textures or maybe doing a Perlin noise mix of two similar textures (e.g. two grass textures, two rock textures, etc). Add a simple water shader. Maybe a better lighting model and selectable skybox.
5. If you're feeling adventurous, add vegetation support or the ability to have voids in the terrain (caves, arches, etc). I know this would be challenging.
6. Can your tool be integrated into somebody's game to allow infinite terrain? This would be a huge plus.

The video also gave the impression that the generated terrains would be quite homogenous, e.g. many shots of infinite mountains, infinite green hills, infinite dunes. This is both a texturing thing and a terrain style thing. I realise that you can use some sort of brush tool to raise or lower an area. I have no idea if I'm a typical user, but I would love the ability to set very general generation parameters by region, e.g. draw out "this circle will be mountainous, this curve will be sand dunes, this curve will be green fields, the rest will be ocean" and the engine would just generate and blend the edges together nicely, rather than bothering myself with raising individual hills.

Maybe more feedback than you need or want, but I hope it helps.

1. A lot of it didn't look FPS to me, perhaps it's just a scaling issue. It felt like you were either flying or a giant in the terrain. That makes sense as a designer, but it might be nice to have a viewing mode as well to get a feel for your terrain. Or just get that feel from point 2.
2. I don't have massive engine experience... I do find Unity quite user friendly, and they have a nice Asset Store which you can usually pick up some cheap or free models/textures from. They also provide some nice demo projects.
3. I think it's in the details... show more examples where you've modified the terrain appropriately, set an appropriate lightbox, etc. Have a particular reference photo in mind and just see how close you can get it (the general feel, not the terrain specifics).
4. Textures etc may be a tricky thing for you. For general game developers there are plenty of sources of royalty-free textures, e.g. the Unity Asset Store, a thousand websites which I use occasionally but can't recall the details of. However you may be in the position of redistribution for commercial purposes because your users may use your textures in their games. I'd be careful on the legalities of that one. You could probably be much freer in your videos, and in-editor if they can't export your textures (less than ideal). For water... I'm not sure. Export it as a simple textured plane, let them apply their own shader? Maybe you could try an SSAO shader, but it's not my area of expertise. With skyboxes you're in a similar position to the textures. It all depends whether you intend to export the skybox. If not, any royalty-free texture set should do.
5. I would never suggest you compete with SpeedTree. ;) Either use an existing solution in your editor, or just use some existing solution for a demo video. Re voids, it depends. I think some engines are happy with just exporting as a mesh with LOD, but I am not an expert.
6. It depends how general your algorithm is. Potentially it could be turned into plugins or similar for other systems.

Your takeaway item: I am not an expert.

Edit: There are tools for making photos tileable, and then you would own the copyright on it. You could create your own textures of grass etc if you're willing to do the legwork. I'm sure the photograph side of it is a little technical, e.g. "from a stepladder pointing straight down at mid-day on an overcast day with no flash".

Also maybe consider using something like parallax occlusion mapping shaders, it allows you to add low-level detail with little detail. That unfortunately requires fancier textures (with heights and normals I believe). Check out one of my old posts to see an example, which involved a fancier variant called cone step mapping. The rocks and wood are all done with textures. CSM is probably overkill (plus I never figured a good source of textures for it), but parallax occlusion would be fine.