1) They really should place two value for these monsters. One for with summons and one without.
Should they?
The Lich in the playtest has two fifth-level spell slots. It can cast Cloudkill.
In quite a number of tactical settings, that is a horrifically powerful spell - 6d8 poison damage every round, for up to ten rounds (as long as the Lich concentrates - and since it doesn't affect undead, the Lich can stay in the cloud happily). The lich is a pretty good tough solo encounter for a party of four 5th-level characters (I know, I ran it as that a couple of weeks ago).
But what if I didn't cast it? I mean, in the fight I'm thinking about, a 5th-level lightning bolt was better for the first round. In some cases, a fifth level magic missile or fireball might be better - they can hit a vastly wider-spread group than a 20' radius area.
But does this mean that the Bestiary should contain two XP values for the lich - one casting Cloudkill, one not?
Of course not.
The XP value needs to show the balanced *potential* value of defeating that creature. If it doesn't use all of its abilities, it's still been defeated, the party just had a (possibly) easier time of it. Maybe the tactical situation was such that it couldn't use those abilities. In that case, hopefully the party had some part in causing the situation that "weakened" the creature. If so, they earned the XP that way. If they had nothing to do with the situation, then the GM has constructed it.
I can see basically three reasons why a GM might weaken a creature by using its full abilities:
1) The GM has a specific goal and is twisting the creature to fit it. This is, I believe (am I wrong?) fairly rare (as in: probably most of us have done it from time to time, but the vast majority of monsters we run are as standard). Some GMs are experienced enough to do this well and can work out a reasonable XP value for the resulting creature. Some GMs are not, and they would probably be best served by looking for a closer match in the bestiary and using that, or looking for another creature from a prior edition and converting it (I've seen some good conversion guides - the 4e one (
http://www.enworld.org/forum/showth...E-edition-monsters-to-D-amp-D-Next-on-the-fly) seems to work really well, which is a great relief for my two 4e games that are converting to 5th)
2) The GM forgot about them. In which case, it seems a shame to penalise the players for a mistake on the GM's part.
3) The GM made a mistake, used a creature that's just too powerful and is trying to tone it down to avoid a TPK (I'm assuming that most of us have been there at some point or other). I think in this case, it's usually better to *not* tone down the creature - to try to come up with something that lets the players still defeat it. Maybe some clever use of scenery, or some allies, or a third side entering the fight. Either way, though, it's an issue when the players defeat it. I would probably search for something close enough in abilities at the right level and switch to that, either during the combat if there was time or semi-retroactively (basically, give the players a clue that the creature is weaker than the thing it looks like, and then deal with what it *really* is when the situation has resolved). For XP, I'd look at how tough the fight was retrospectively and then set a reasonable value for the fight based on the number of players and difficulty - effectively, encounter design in reverse.
... that tangented more than I expected. Oh well.
XP for summoned monsters is factored into the summoner's stats.
Immunity to SA or magic "needs to be used judiciously". I take this as a coded "Sneak Attack will affect all creatures" (but "Turn undead" will only affect undead). Feels lazy/uninspired to me. 
There will be a spell point option.
