If I understood correctly, when I add an Aux Send to a bus/track, I have to adjust the mixer strip of that channel and the send gain of the Aux to adjust the levels of the clean and "auxed" signal respectively. I'm not sure how this works in other DAWs, but isn't it a bit counter-intuitive? One would expect that the mixer strip adjusts the global volume of that bus, no matter what effects one applies to it, even in Sends - instead as it is now it could be all turned down, but if some part of the signal is going through Aux it will still be heard. Mixing also becomes much more difficult: every time you want to adjust the overall volume of a track that has n Auxs, you have to move n+1 sliders instead of 1.
Wouldn't it be more obvious to have two sliders in the Aux Send window, one for send gain and one for dry/wet proportion, exactly like in the Insert one?
Also, another concern of the current architecture of Auxs. If I get it right, putting two Auxs in a row, your signal will go through these in parallel, not in series.. So you couldn't have a distortion bus, a delay bus, and then add two Auxs to a track to have a distorted delay.
All in all as I see it - couldn't the Aux Send be a perfect replication of the Insert pseudo-plugin, with the only added benefit of having a drop down menu to choose the bus, without having to manually define sends and returns (i.e. sends would automatically be connected to the Ins of the selected Bus, and returns connected to the Outs) ?