This could help solve two problems:
1) In some Italian 2/1-like systems 3-level jump reverses e.g. of the form
1x-1y
3z*,
* y>z>x
shows 6+x5z and a hand too weak for a normal reverse.
I believe this is a much better use of the jump reverse than mini-Splinter, but unfortunately the bidding is sometimes already too high.
It seems like Responder would be able to pass 3z more often if it showed 5(+)x6+z instead.
2) In my (range-first) relay structure I currently resolve (6511) at 4♦ (more than low enough IMO) but (6520) at 4♠ (too high).
A solution, after a 1x opening with 5x6z, could be to treat
5x6z20 (currently resolved at 4♠)
as equivalent to (a very dissimilar shape such as) e.g.
6x0z34 (currently resolved at 4♣)
when resolving shape, but rely on Asker's ability to "guess" (by looking at his own hand and drawing conclusions from opps' silence) which of these two shapes Teller actually has.
(Guessing whether Teller has e.g.
6x5z20
or
6x0z34,
after a 1x opening (and no canapé with this shape), could be significantly more difficult.)