Yes I have the same render target being used, it is just that it takes two pass's for both the graphical data and depth data to be used as when you set writeDepth="true" this also results in it using the depth buffering when rendering.
Code:
<Stage id="Translucent">
<SwitchTarget target="PARTICLES" />
<ClearTarget depthBuf="true" colBuf0="true" />
<DrawGeometry context="TRANSLUCENT" class="Translucent" />
<DrawGeometry context="TRANSLUCENT_DEPTH" class="Translucent" />
</Stage>
is what my pipeline looks like at present, the TRANSLUCENT context is the same as from the examples and the TRANSLUCENT_DEPTH simply does nothing in the fragment shader but has writeDepth="true". This render context then gets combined with the screen buffer using a fragment shader to test its depth values against the GBUFFER. This works but seems wasteful having to 'render' all the particles twice