Below is a outline of how I typically do my own forecasts for my personal flying.  This approach considers just the basics - for certain conditions an amended approach would be used, or someone forecasting for a contest or a very long flight might want to spend more time than I do, etc..  Note that (1) I fly a sailplane, which biases my forecasting toward sailplane cross-country (e.g. through my choice of a minimum B/S ratio), (2) clouds are usually not a factor for my location, so their prediction is not a large concern in my methodology (in fact I often skip the cloud forecast parameters), and (3) I normally fly over the higher terrain near my airport, so am not usually concerned with "flatland" conditions. 

• I look at the "Height of Critical Updraft Strength (Hcrit)" forecast just after mid-day (solar noon) to see if thermals will take me high enough to undertake the flight I contemplate - if not, then I need not look at other parameters.  Caveat: If other parameters could affect the thermalling height I need -- such as lift from convective clouds or convergence -- those parameters should also be looked at. 
• I look at the "COMPOSITE: Thermal Updraft Velocity & Buoyancy/Shear (B/S) Ratio" forecast to see if the thermals produced will be workable, i.e. a B/S Ratio above 4 and even better above 7 - if not, then I need go no further.  Caveats: (1) the B/S value below which thermal are not workable depends upon glider type and pilot experience,  (2) If convective clouds produce significant "cloud suck", then B/S and W* forecast values are underestimates since they do not include that cloud effect,  (3) I am not much concerned with "Thermal Updraft Velocity (W*)" per se because it is included in the Hcrit calculation and because I fly where terrain is significant (i.e. not under flatland conditions) and know that if I can reach my desired Hcrit in that terrain then I will be OK. 
• If clouds might be important then I look at the "COMPOSITE: Cu Cloudbase for Cu Potential > 0" forecast to see if clouds might prevent me from reaching my desired thermalling height - but in addition, the presence of clouds might produce additional "cloud suck", which is neglected in the computation of "dry thermal" parameters such as Hcrit and B/S Ratio and W* and so can alter expectations based on those parameters.  If much cloudiness is indicated, then I might also check out the possibility of overdevelopment using the "COMPOSITE: OD Cloudbase for OD Potential > 0" forecast. 
• I look at the "COMPOSITE: BL Wind Speed and Direction" forecast. 
• I look at any locally important parameter - since I am flying around mountains, one such is the "convergence" forecast ("BL Max.Up/Down Motion"). 
• I look at the local weather service forecast for my area, looking for factors beyond these soaring parameters which their forecasters might highlight, notably the chance of rain, thunderstorms, etc - such factors are best determined by experience rather from a simple model output.  (Even better can be the "Forecast Discussion" which every NWS regional forecast office provides - while somewhat technical, they provide valuable insight into what experienced meteorologists think of the current conditions and forecasts.)
• I look at a detailed resolution (1 km) visible satellite photo to see if there are cloud factors which the models might have missed, such as high cirrus (which is often missed by model forecasts but which can greatly diminish thermal strengths). 
• Occasionally I will do a more detailed analysis, e.g. looking at other parameters, different times, comparison of forecasts from different models, looking at forecast soundings, etc. 

Additional Information:
BASIC Thermal Forecast Parameters
Parameter Descriptions
Additional helpful information - see "For NEW USERS" and "What NEXT?" sections