Your idea of wiping at the end of the X axis end up being a more practical solution. I wanted to use water soluble PVA for support material. But it is soft, and does not stick well to the base. Makes a huge mess when the wiping tower topples over.
Here is the revised design. .5mm plastic sheet wrapped 3 times with kapton and a small hole drilled in the middle. During tool change, retiring nozzle is retracted couple of mm. New nozzle is positioned over the small hole, and primed for a few mm length. Small hole keeps the filament from sticking to the nozzle as is priming, and creating a big mess again. Once priming is done, nozzle moves away from the hole, and the length of filament is broken off the nozzle. Now the nozzle is ready to build the part.
Had to put the wiping of both nozzle on the right side. Can't do right nozzle wipe on right and left nozzle wipe on left, because that would require an "if" statement in the tool change G-code. Searched all over the net, and can't find if "if" statement that is supported in the tool change G-code. Only placeholder variable was supported, and that was not enough.
Result was pretty good. No blobs or ooze. PVA itself was tricky to use. That is a different topic. Had to build a solid layer of PVA below the ABS plastic for the ABS to stick on top of the PVA. Once that is done, test results looks great. Now on to building some real parts...
Put all the files here for anyone else who wants to experiment:http://benmlee.com/Bukobot-Wipe.htm