Well its basically same as not using --jobtimeout at all. Unless you use small values(like 0-120 or so) that it happens sooner - so they work. Really high values seem not to work though.
Just read that --jobtimeout 300 is the default and that decreasing n below 80-100 may cause instabilities, depends also on the pool. Out of curiosity what is your ping to the pool and did you try if another pool gives the same issue?