Any studies that have shown any indication that grounded representations might actually be something?

Witt et al., 2010 found that Object naming (classification) while participants squeezed and maintained constant pressure on a small foam ball with either left or right hand found that hand related items took longer to name. Why? Maybe because the brain area is already in use, indicating a bottleneck. Another study found the same with both hand related and foot related action verbs while classifying items and checking for response time.