Hi, This will be a good one for Nick to answer but I can tell you what I have found in practice. It is possible to get a different result when you have a mark left to one side in the middle of a route. In general, I find the routings stay further away from the new mark than in the absence of that mark. I find that the use of Race Notes over land and areas you are trying to avoid works better than additional waypoints. So as a general rule use as few waypoints as possible. In the situation you are describing it can be instructive to look at the reverse and normal isochones and to shade the routing (with different times eg 15 min then 60 min) to see what is going on. It can be that two very different routes have similar times. If you are using a current GRIB, stop using this to see if this changes the routings. Hope this helps.
Cheers, Will.

Good question!
Here's my guess....
If the isochrones were perfect circles the opts would probably be the same. That would only occur if your polars were only 1 speed no matter what the wind was doing.

As soon as you plug various polar speeds and various winds in, the isochrones are no longer perfect circles So the isochrones and the reverse isochrones would no longer meet at the center of the course. Normal circumstance

When you put in a third midmark the program points the isochronal circle toward that mark even though the program 'knows' you don't have to go 'to' it. (Pointing the polar at the mark would be the normal way of doing it.)  but because the isochrones are not perfect circles, the polar driven shape when pointed at Mark 2 vs. Mark 3 give a different solution.

As Will mentioned, the better way in this situation would be using the Race Note to make sure the Opt went on the water side of Spain!

That's my story and I'm stickin' to it (For you Yanks)
Cheers,
Ernie