There shouldn't be a size limit on log files to scan, since we scan and check one line at a time and free the used memory in between.
Some performance tests were done with text files, each with about 11MB of data in them. The lines of text did not exceed 150 characters. These files each took about 0.5 sec to scan, and this should be fairly linear for smaller/bigger files.
So 300-400 files of this size would most likely take around 3 minutes to process. If each file is 100MB it would take closer to 30 minutes. If each file is 1MB it would take 15-20 sec.
Another factor to note is the length of each line. Longer lines would need less line scans, but longer time to process each line and opposite for shorter lines.
Ultimately, the only limitation is the amount of time taken and there is no limitation on raw file size.