This goes to what I have always said about the dangers of a GM going into a draft with his mind made up that he has to do one thing or shouldn't do another. GMs need to be flexible. Just because you have hit on late round picks in the past doesn't mean you should pass on superior talent when it is there just because you feel you need to draft another position.It just depends on the player, as well as the other players left on the board at that time.
Early OL tend to be among the safer positional picks, so if you have an OL rated highly beside players at other positions with question marks, take the OL.
You must have missed the public service announcement from the CECICIMB in the AJ Dillon thread. I'll try to give you the short version. If you use words like " chance" or "tend to" you must make sure you also acknowledge the possibility that the opposite may occur otherwise someone will be forced to point it out. Even though words like "chance" and "tend to" come with a built in implication that the outcome may not happen.They tend to be... doesn't mean they are.
Always depends on the individual.
For example if you say early OL picks tend to be safer picks without specifically mentioning that it is not always the case someone will be forced to bring up Tony Mandarich.
For the record, that year I wanted Thomas or Sanders. Wasn't happy when they went with Mamwich.
Lots of people would have preferred another player. Of course most of those people came to that decision a year or two after the draft. While Sanders or Thomas would have been great picks I think you would be hard pressed to find many people who thought Mandarich was not a top 5 talent.