Well, I can assure you that converting to Vertex Buffers will improve the use of bandwidth when sending data to the graphics chip and the rate at which the card is able to fetch that data, but...

If that's not where you're bottleneck is, then it's not surprising that you're not seeing an improvement in framerate. Considering the age of your card, I would be willing to hazard a guess that you may be highly fillrate bound, especially if you are using any sort of multipass/multitexture operations. Try this: Generate all the landscape data as usual (upload vertex data, set textures, etc.) but leave out the actual render call (in this case glDrawArrays or glDrawElements) If this gives you a sudden FPS hike, then you're almost certainly fillrate bound. The other scenario (though I feel like it's unlikely) is that you're CPU bound, in which case you'd have to optimize your game logic to see any improvement.

One final idea: It may be that your card, as old as it is, doesn't have the proper hardware support for vertex buffers and as a result is emulating them, storing the vertex data in system memory instead of video memory. (Which would actually make this a form of CPU bottleneck) If this is the case then it's unlikely that your system will show any benefits from the conversion, but anyone who's card supports the method will enjoy a speed boost.

Quote:I made a heightmap using vertex buffers (thanks a lot for a lot of you!) but I see no improvments in the FTP from my older Display list method.

u wont VBO + display lists are normally about the same speed BUT as Toji said the bottleneck is most likely elsewhere

Believe it or not, drawing lines is always slower than drawing filled triangles. Another way to test for fillrate limits is to make the window smaller. If the framerate increases, then you are fillrate bound.