John,
A couple of issues are raised here, firstly - and probably most significant is that I don't think a pump from an external filter would give you sufficient pressure to efficiently run your R/O unit I have to have a dedicated pump to effectively run mine and that pushes through water at around 75psi (don't know what that is in bars, but it's quite high). Were you to buy a pump then you could try your theory, but bearing in mind that well over 80% of water passing over an R/O membrane goes to waste you would still find yourself using a lot of the water to attend to your terrestrial plants - I fear.
Now it is possible that the rejection rate of 'used' tank water will be less than from straight tap water, but that's another theory which would need testing...
But considering that the lower the pressure the water is passing over the membrane at the higher the rejection percentage becomes.
I'd suggest, if you have the R/O unit then give it a try...you've little to lose, it shouldn't damage the filter pump if you find that it isn't able to build up the required pressure - so long as you don't leave it too long 'trying'. You would, like you said in your original post, need to replace the lost trace elements too.

Some years ago I built a bit of a makeshift filtration plant using an old Eheim filter body with floss in the lower part and Zeolite above, I would syphon all my waste water through this and reuse a percentage of the resultant water mixed with fresh tapwater to replace the waste I had syphoned out of the tanks...the Zeolite could be 'revitalised' with cooking salt and the floss could be washed out and then both could be reused at the next water change.
I never was able to form any concrete conclusions as to how effective this actually was, and eventually it just became too much hassle to continue...but no fish were lost during the experiement period, it might be worth re-enacting this in these days of potential water metering...I might look into that.

John