In the recent decades, there has been growing interest in novel applications of neuro-rehabilitation and motor power augmentation based on the experimental paradigm of Motor Imagery (MI) and Motor Execution (ME) respectively. Qualitative comparison between MI and ME reveals significant difference in response time, activation magnitude, and cortical source distribution. Based on these differences, different frameworks of MI and ME have been developed for enhanced Brain Computer Interface (BCI) applications by providing task specific information in time and cortical source space. Decoding of subject's intent using cortical source domain processing increases the accuracy in the MI and ME scenario by 5.18% and 3.68% respectively, compared to conventional sensor domain processing. The performance of the proposed methodology outperforms the stat-of-the-art techniques. The present study highlights the limitation of cross framework i.e., utilization of MI framework in ME experimental paradigm or vice versa. In particular, the performance degradation of MI and ME tasks was observed in the cross framework scenario. The experimental results are expected to provide an evidence for researchers when designing an imagery or execution based online (or offline) BCI system.