For ``small'' systems like the one we benchmarked the solvers on here, it seems that OPTIMQR has a potential to run a lot faster than ``smarter'' solvers that utilize more sophisticated pivoting strategies.
It is uncertain however, how well OPTIMQR would perform give a much larger system. The QR factorization is inherently slower than LU and many other factorizations. When the savings from not pivoting rows are swallowed by the price of using a more expensive factorization, is still unknown.
To be fair, it should be mentioned, that the SUPERLU package can solve may kinds of systems, not necessarily just systems with fixed sparsity patterns (although SUPERLU does benefit from this).
If one needs to solve a sparse system of linear equations, where the sparsity pattern is fixed, the OPTIMQR package should definitely be tried out. Spending a few hours having a good solver generated, can earn them selves many times, on some systems at least.