Does anyone know of an app (or any purchase-able software), that can solve arbitrary, user-created life and death problems?

( I see this was already a solved problem in chess software over 10 years ago... )

'arbitrary' - that is the difficult part.

Many years ago I got to a know a Chinese guy on the Internet who wrote a program that could solve with very high degree of success any 'closed' L&D problems (the problem has to be confined in a closed area so that brutal force type of algorithm can work). By that time there were not many good AI codes available so that was quite a revolutionary code. But somehow this person along with his program disappeared. I still don't know why. That could have been a very successful product.

The problem with many L&D programs is that the solutions, often prepared manually, cannot possibly cover all the variations. A strong player can see that some points are clearly not working. But weaker players, quite legitimately, want to try those 'clearly incorrect' moves and programs often cannot respond to those move. Some form of AI can fill the gap nicely.

GoTools comes to mind. I have never used it, however. http://lie.ac.brocku.ca/gotools/

_________________
Occupy Babel!

SmartGo for Windows embeds gotools. You may have to add a number of extra stones around a problem to make the boundary clear. I stopped trusting it years ago, eventually concluding that it was more important to use my brain than to know the "right" answer.

There used to be something called GoGrinder. But the "arbitrary" adjective is the crux. That would seem to be equivalent to solving go.

Even a small problem can end up needing a 9x9 sub-board to display. So, yes, the ability to solve any problem would appear to be the same as solving 9x9, which is still unsolved.

_________________
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).

I'd say that in chess things are not as solved as you think. There are still chess problems that computers are not great at/unable to solve. These usually involve some mix of blocked positions, fortresses, or long move horizons.

A couple of examples that come to mind are the forced mate in 262 (6k1/5n2/8/8/8/5n2/1RK5/1N6 w - - 0 1), The Behting Study (8/8/7p/3KNN1k/2p4p/8/3P2p1/8 w - - 0 1), and a legal position that crashes some modern chess engines out of Leonid's mate collection (brqqqqqk/B6q/R6q/Q2Nn2q/Q2nN2q/Q6r/Q6b/KQQQQQRB w - -).

You are right that for most problems chess engines are excellent.