Good info from one so his comments here.
While not necessarily always an improvement, switching out soldiers amulet for knights amulet gives the build some interesting changes which seem to improve the build in the majority of situations. Because the mitigation from your base armor is equal, we can ignore it to compare the effective health changes. The mitigation from toughness is equal to 1-(BAR/(BAR+TOUGH) where BAR is your base armor and TOUGH is your toughness stat, and your effective health for this comparison is simply 1+(1-(BAR/(BAR+TOUGH)) times your health.
With the soldiers amulet, we get 1+(1-(2127/(2127+1950) * 17945 = 26528 effective health.
With the knights amulet, we get 1+(1-(2127/(2127+1381) * 20235 = 28201 effective health.
When looking at this, you have to both consider that toughness makes the health you restore though your heals harder to take away, meaning your heals with the soldiers amulet are roughly 8.5% stronger [which is from the difference in the mitigation from the toughness, or (1-(2127/(2127+1950)))-(1-(2127/(2127+1381)))]. What this means is that to make up the 1673 difference in the effective health in favor of the knights amulet, you would have to gain that 1673 health back through healing. It takes 1673/.085 to gain that back, meaning you need to heal/be healed for 19682 health to make that up. This is a tough value to reach, but in long fights does mean the soldiers amulet would be stronger.
Of course, this is all assuming that all the damage you are taking is not condition damage, which totally ignores the toughness and only takes your raw health into account, meaning that if you aren’t able to keep conditions off you effectively (which is not always possible despite it being a strength of this build) the knights amulet becomes a better choice.
The other stat changes are the power change and the critical strike chance change. The net change is a loss of 229 power and a gain of 27.09% crit chance. To examine how these change affect damage output, we will first look at how power affects abilities.
When calculating damage of an ability, the total power of your character is multiplied by a static power coefficient and a static number based on your weapon damage. Most power coefficients are low, usually around .3 and almost always under .5 with channeled abilities being the exception, with lower damaging abilities having lower coefficients and higher ones having higher coefficients. This means that on most of your abilities the most damage you will be losing will be about 115 damage (assuming a .5 coefficient) and on many you will only be losing about 68 damage (assuming a .3 coefficient).
For the crit chance change, we simply multiply the change in crit chance, 27.09% by the critical damage multiplier (in both builds a base of .5 + .35 from talents, equalling a bonus 85% damage). This gives us a 23.03% damage increase over the other build, which should, especially in combination with the sigils of superior blood and earth, which mean additional damage increases from the bonus 27.1% crit chance (.271*.3*325 = 26.42 extra damage/healing on each attack on average from the sigil of superior blood [the 325 damage/healing on proc is fromhttp://www.guildwars2guru.com/topic/36865-detailed-effects-of-sigils/%5Dand, assuming the 5s of bleed rounds up to 6 from the runes, 6*(40+300/20)*.6*.271 = 53.66 extra damage on each attack from the sigil of superior earth).
I think that this shows that the knights amulet is better in terms of survivability and damage output in almost all scenarios (with the soldiers amulet being the stronger choice in long sustained fights with little condition damage), although if someone notices a flaw in my math or something I have overlooked, please let me know. I’m running through similar scenarios with my thief build, and would like to know both for myself and for other people who read this/my builds if I’m doing something wrong.