Steady Shot and Multi Shot are both physical damage, and would benefit from the armor reduction. Only Arcane Shot (and Serpent Sting) would be unaffected.

Are you getting your Kill Commands in there and are they adding any delay at that speed?

I see this a lot and I think it is worth noting that we currently don't know exactly what the latency shown by the WoW client is. If it is the standard use of the word then it is the round-trip-time of a command to the server. Time from when I send the server a command, till the server sends me back a response responding to that command. This includes time for the command to traverse the internet, be processed and acknowledged. I'm pretty sure this is the case since when the server is overloaded latency shoots up. This has nothing to do with the internet itself, just the limited processing power of the server, it cannot do all the things it needs to (including responding to incoming commands).

If that is correct then a castsequence macro won't incur a double latency penalty.

Example with 200ms latency (round-trip-time), 90ms one way travel time

000ms - press button
005ms - command sent to server
095ms - server gets command and queues it
105ms - server processes command
110ms - server sends response
200ms - client gets response

So anything that is processed entirely on the server, you should incur half of your total latency to get the results.

000ms - auto shot timer finishes
005ms - server processes auto shot
010ms - server sends results
100ms - client gets results

If you want to perform an action immediately after that, you will incur your latency + reaction/macro mashing delay. If I am hitting it 10 times a second (100ms macro delay).

200ms - client hits macro
205ms - command sent to server
295ms - server gets command and queues it

I think you're correct in that I'm being overly simplistic in my evaluation of effective average RAP on that trinket. However, I'm not sure I agree with your alternative methodology. If you wanted to be very specific (which is what I'm gathering from what you've listed), you would have to know what the exact breakdown of the components of your dps. You've listed autoshot at 50%, but presumably that leaves steady shot as the other 50% of your personal dps (completely ignoring pets). Since steady shot gets 0.2*RAP added as the RAP contribution, you'd have to have 0.5*7.5/0.2=18.75 RAP for steady shot, which would add in to the total.

Honestly, you'd be better off just using Cheeky's spreadsheet and comparing your personal damage gain by manually decreasing the boss mitigation by 300 and then muliplying that by the uptime to get a rough estimate of added dps. Honestly, the exact result is purely academic for the question of preferential use of trinkets; apparently without respect to the method we chose to use Madness of the Betrayer outpaces the alternative trinkets, which was the original question.

I acknowledge that I frequently simplify calculations to what some would deem an inappropriate level, but in cases where I'm speaking generally I will often make basic assumptions (such as assuming all damage is mitigated equally, which isn't true - Arcane shot being an example of damage that is unmitigated), but I'll try and call those out when I make them such that they don't cause confusion.

Latency aside, how does machine performance factor in? For example, would it impact your macro-mashing raid performance to be getting 10 fps?

A good read for some people that use Macros for shots and are getting low FPS.
2.1 - Macros Directly affecting fps?

Edit: Fixed url.

Another thing furo, is i believe KC delays your auto if fired in the .5 second cast time of the auto. Because of this I don't care for auto/steady with KC added in macros.

I've got a question, do you hunters ever use amplify/dampen magic on your pets for boss encounters? I've tested amp. against mend pet and it doesn't affect it, id assume the same for dampen. I can see a use for dampen on fights with a lot of magic aoe (Void Reaver, Fathom-Lord, possibly Lurker) where it would reduce the damage taken, however the other side of the coin is the -effect on other heals my pet receives (chain heal, PoM, PoH and 2 piece t5).

I'll test out damp if we do Void Reaver tonight and post results. I generally let him sit in through pounding, since he's get max AR.

Just like any other special it does delay autoshots.

Post from Survival 2.1 raiding just in case any helpful soul reads this topic only. I guess it can be posted here to since its general stuff.
SURVIVAL Raiding in 2.1

Its few pieces of my combatlog from yesterday Vashj. And there are strange things there that I can't explain. (or can't read combatlog properly)

Like questions:
*Does multi shot really delays auto shot? I clipped auto, but still auto hit the target 15ms after multi. While some clipped autos seem to be ok with 0.5s auto shot delay after multi.

*Where did i get 200 ms faster attack speed of the blue once?

Went checking logs for crazy clock and its not there. All 3 logs match, wws doesnt report crazy clock either during merge.

And what the *. Missed EW?!

The more that I'm looking at log more crazy its getting.
Its driving me nuts :/