While, Reaction time (RT) studies have been in vogue for a long time, Donders was the first to try and get to extract processing time (time taken to identify the target) in the 1880s. The subtraction approach however has a number of limitations - there are multiple covert factors at play, high variability and speed-accuracy trade-offs that take place. Also visual stimuli tend to leave a after-image so the timing is unpredictable. To counter this people have tried using mask images but that introduces other confounds.
Shwetha's group used a compelled-saccade task with a variable gap time between go and cue images to arrive at processing time (reaction time - gap time). They also created a Tachometric curve (with centre point, slope/rise time and maximum height) that graphed %correct responses as a function of processing time. This approach yields much more information than reaction time or psychometric curve (%correct vs gap time) studies.
They manipulated three factors (motivation; stimulus discrimination and features; perceptual learning) that affect perception to see how the tachometric curve responds.
Areas that need further work include effect of expectation, multisensory decision making and neural correlates of perceptual changes (areas and mechanisms).