It depends on which type of scale is the most important for your research.
I presume you're are talking about geometric scale. Google Earth has an imagery with geometric resolution of 1.5 meters, and Landsat's resolution is 30 meters. So, Google Earth is better.
But, if the time span is also important, Landsat imagery is far better because it gives historical data for almost 40 years, and Google imagery gives only 10 years.
Landsat has also much better spectral resolution than Google Earth, which is important for land cover classification.
I agree with Tomislav's comments. With Google Earth imagery, the level of details which you can see are definitely unparalleled because of the very high spatial resolutions of satellite images which form the base for Google Earth Images. I generally use Google Earth Imagery for visualization and to get a 3D perspective in parallel with pan-sharpened Landsat Imagery (15m) of the study area. Therefore for quantitative analysis I use satellite images such as Landsat (15-30-100m) and ASTER (15-30-90m) combined with qualitative studies of the same area on Google Earth. The Spectral, Radiometric and Temporal resolutions of Landsat images surpasses that of Google Earth images.
When you mention 'fine scale', what are the dimensions of this fine scale? What is the minimum size of the unit on the ground that you want to study? I would suggest you to first make a list of the characteristics and properties of the object on the ground which in your case is woody vegetation. Once you identify the properties of the object under study you can then look for the best monitoring tool based on the Spatial, Spectral, Radiometric and Temporal coverages of satellites, airborne and UAV sensors. I have also attached an article which discusses Google Earth imagery.
When using Google Earth images over time, be careful, because the images from earlier versions may not align properly if overlain on contemporary images of the same location.