SAM (version v4.0) is free and a software very intuitive. SAM reports the residuals. I like.
ArcGIS implements GWR Software. ArcGIS does good results reports.
GWR 4.0 (free). This is a specific software. It has some limitations in relation to data import and reports. However, GWR includes logistic distribution to binomial data. GWR was continued in R.
SpaceStat. May be the most complete software package (not counting R) for spatial analysis of vector data. I tested the trial version.
Raster / vector data
Saga GIS. You can fit the GWR using raster and vector data.
Grass GIS. the "r.gwr" runs in the same way.
R. Here, you can use several packages: GWmodel, gwrr, spgwr
I agree. I also recomend R. However, it need some experience. At first look, is less intuitive, but as we progress, we increase the productivity and the efficiency by using R. Nowdays, I use R in almost everything. Now, the authors of GWR (4.0) develop the method in several packages for R. This is yet another reason why I use R for this purpose, and others..
As suggested above, the choice of software depends on your level of experience using the programs. If you're comfortable with R, I think that will be your best bet. Alternatively, if you are looking to learn how to use a new program specifically for GWR, I will recommend that you go with GWR (4.0). It's a point-and-click, easy to learn program for GWR. Not to mention that it's free. I hope this helps