Actually, when it comes to watchability, refs and SotG are about the same. What determines watchability is the execution. We can all remember ultimate games that were awful and painful to watch and we can all remember basketball/football/soccer games that were awful and painful to watch. There are certain things that make a game more watchable and more exciting and things that detract from that, but reffing and SotG don't really have anything to do with those.
First off, no one likes to watch a blow out. The other day I went to watch the (evil) Cardinal play the Lady Ducks and it sucked. Other than Nia Jackson's valiant drives it was a depressing display. There's only so many uncontested offensive rebounds followed by uncontested lay-ups you can watch. There's only so much interest in wondering if the Ducks can hold the lead under 20...30...40... But refs and SotG don't have anything to do with that. Blow outs happen.
No one like to watch a game with a ton of stoppages. I mildly follow high school basketball here in western Oregon and twice recently, teams have used the hold-the-ball technique to eat the clock. There isn't a shot clock, so if you are undefended, you can hold it indefinitely. In the recent Mapleton-Triangle Lake game, Mapleton held the ball (literally held it) for 2 minutes of the 4th quarter while clinging to a four-point lead. A handful of years ago, Siuslaw held the ball the entire first quarter, before putting up a three at the buzzer. First quarter score? 0-0. There are always nooks and crannies in the rules for teams to exploit - if they are willing.
While there is no difference here between a reffed and SotG in theory, although there are some much needed changes that need to be made to the observing system to effect the necessary speeding up of the game. A couple of years ago at Solstice, I implemented an observing system designed to speed the game up and make it more watchable. I went through a ton of uncut game footage and timed and categorized all the stoppages. Here were the biggest issues: time between points, bricks, arguments, calls. Most frequent call resulting in stoppages? Foul on the mark. Traveling was not a major issue. (Footage was from club nationals.) Here are the important changes to remedy these issues.
1. Less time between points. The NFL runs 40 seconds. Surely we can get to 60. It will be an adjustment, but not nearly the adjustment going to 90 seconds was. In the 90s, Schwa (masters of eating the clock) once were timed at 5 minutes between points!
2. Brick at midfield. A brick almost doubles the time between points as Joe Handler swaggers off to get the disc, survey the field, check it in....yawn. Defenders hated this rule, but the number of bricks dropped precipitously when we tried this at Solstice.
3. Quick observer rulings. The current policy of time for players discussion of calls achieves two bad things, while failing to achieve its stated goal of allowing for player determination of a call. First of all, letting the players discuss means any call that needs discussion takes 2 or 3 or more minutes to resolve before play can get going again. Also, it creates an opportunity for drama. When observers step in quickly, the opportunity for players to argue and carry on is snuffed out. The supposed goal of observers waiting is for the players to have time to resolve the call, but that only happens maybe 5% of the time and the remaining 95% of calls result in argument, drama and an observer ruling. At Solstice, we used a simple "walk-to" rule. When a call was made, we (the observers) started walking to the play. When we got there, we asked the players, "do you have a call?" If they said yes, they made it. If they were still arguing, we made the call.
4. TMFs for bad calls. The object of this policy is to prevent teams from controlling the game by making lots and lots of calls. Once it is clear that a team is going to make lots and lots of calls, they should be held to a high standard of accuracy and punished if they don't reach it. Also, players should get TMFs for cheating even if it isn't called. At the club level, only about 10% of marking fouls are called because it is such a disadvantage to stop play. These uncalled fouls should generate TMFs even if the thrower isn't calling them.