Physical coverage plays a pivotal role in detecting mis-assembled regions: even if the sequence coverage track shows no interruptions, its possible that the physical coverage one will, thus hinting at a possible chimeric sequence:
An example of a chimeric contig identified plotting the physical coverage track of a mate paired library.
The above example, from my PhD thesis, was obtained in the early days of NGS when mixing technologies (Roche 454 for producing contigs, thanks to the longest reads and SOLiD Mate Paired library for scaffolding) was as common as painful².