Did some more tests with a fresh simcraft, only things changed were profiles and max_time to 333, so the duration turns out closer to 300seconds:
Warlock_T8_00_56_15 and Warlock_T8_03_52_16 being base ones(with added DE).
- _immo keeps using immolate throughout the whole fight
- _fs uses fire stone instead of spell stone
- _metag uses metamorphosis glyph instead of immolate one
- _immo_fs is immo and fs combined
- _metag_fs is metamorphosis glyph and fire stone
- _metag_fs_immo metamorphosis glyph, fire stone and immo combined

Its all not much, but as it stands now those specs need everything they can get, as my mages would love me if i could keep my deep demo spec after 3.1. :p

This is a 1-part question and 1-part response to this. I too am trying to investigate and justify bringing a deep-demo lock for DP to the raid. We currently run with about 1 SPriest, 2-3 Locks, 2-4 Mages, and a Boomkin at any given point. This means 6-9 caster DPS at any point. I ran the following 2 sims to show the difference between a raid with and without DP:

Without DP:
7004 11.2% Warlock_T8_53_00_18_DG
6936 11.1% Warlock_T8_00_40_31
6689 10.7% Warlock_T8_53_00_18
6689 10.7% Warlock_T8_00_41_30
6486 10.4% Warlock_T8_00_56_15
6083 9.7% Mage_T8_20_51_00
5929 9.5% Druid_T8_58_00_13_SF
5801 9.3% Priest_T8_13_00_58
5586 8.9% Mage_T8_52_19_00
5321 8.5% Shaman_T8_57_14_00

With DP:
7371 12.5% Warlock_T8_53_00_18_DG
7259 12.3% Warlock_T8_00_40_31
7047 11.9% Warlock_T8_53_00_18
7011 11.9% Warlock_T8_00_41_30
6497 11.0% Mage_T8_20_51_00
6263 10.6% Druid_T8_58_00_13_SF
6159 10.4% Priest_T8_13_00_58
5903 10.0% Mage_T8_52_19_00
5627 9.5% Shaman_T8_57_14_00

I simulated giving DP to everyone by increasing the spellpower for all their gear sets by 300 (which I found to be my average DP buff in mostly BIS gear on live). The result is pretty obvious: almost every class gets (on average) 300 or more DPS from 300 spellpower. So really the question for your raid becomes: is 300 spellpower (and even more in Ulduar) beat 280 spellpower from Elemental Shaman?

In my guild we have no Elemental Shaman so it's unquestionable that 300 spellpower as compared to the 165 from Flametongue Totem (135 spellpower increase) across all 6-9 casters easily makes up for the potential 600 dps decrease inherent in the DP spec. However as spellpower increases from gear, it will become even better for the raid, therefore even more justifiable. It looks to be far harder (if not impossible) to justify it if you already have an Elemental Shaman.

In the end it all comes down to your individual raid composition, and for guilds with a "sometime Shaman" it may be beneficial to utilize the dual-spec feature to spec a pure-dps build and a DP build for the times when he's not there.

My question to Zel is: does "optimal raid = 1" include an Elemental Shaman, or just Flametongue Totem?

edit: Answered my own question: it looks like optimal raid includes the +280 spellpower from Elemental Shaman. This doesn't change my results though. Ideally speaking, an Elemental Shaman is about 1,000 dps less than a DP-lock, so if you have a good enough Retadin to cover the 3%, having a Lock go Meta and replacing that Shaman's spot is probably advantageous. But now I'm getting off-topic. :-)

On PTR:
They however appear to hit for the same amount still. I want to look at this because of:
PTR Build 9704 - WotLK Wiki

Suggesting it had been nerfed.

All the set bonuses, and most of the new trinkets, are implemented in SVN, so they'll be in the next release.