I'm researching what is necessary to bring the numeric algorithms into
namespace std::ranges (for those wondering why it's non-trivial: they
require close study like those in <algorithm> got from N3351 + ranges
work). Part of this research involves me consistently scratching my head
over whether or not all of the numeric algorithms should be imported into
std::ranges, or if it's possible to get away with just reduce,
transform_reduce, exclusive_scan, inclusive_scan, transform_exclusive_scan,
transform_inclusive_scan, and adjacent_difference. iota is deliberately not
in that list, but that's also not relevant to the discussion I'm wanting to
have.

Note that I'm only considering the overloads without an ExecutionPolicy.

Are there any practical reasons for reduce(begin(v), end(v), 0, plus{}) to
permit out-of-order computation? For example, is this allowed as an
implementation?

```
template<class InputIterator, class T, class BOp>
T reduce(InputIterator first, InputIterator last, T init, BOp bop)
{
   #if NDEBUG
      return reduce(execution::par_unseq, first, last, init, bop);
   #else
      return accumulate(first, last, init, bop);
   #endif // NDEBUG
}
```

If either question can be answered with "yes", then forgetting accumulate
is a non-starter, and I'll work on nailing down all of the numeric
algorithms. I suspect this is the case, but lack the experience to answer
them myself, and am yet to find a proposal or minutes detailing this
discussion.

If there aren't advantages for this overload of reduce, et al. to co-exist
with their C++98 counterparts, then I'm inclined to propose that only the
C++17 names be ported (plus adjacent_difference), because:

   1. It seems redundant for us to have two algorithms that are specified
   differently, but produce equivalent results.
      1. The ExecutionPolicy numeric algorithms can still have formal
      requirements for associativity and commutativity.
      2. I've had a teacher pass on a question to me from a student asking
   if they should be using accumulate or reduce, so I have teachability
   concerns.
      1. It's likely that there are more than two people in the situation
      of 'Curious Student Reading the C++ Reference' and 'Hesitant Teacher'.
      2. If there's only one name in namespace std::ranges, then advice
      boils down to "prefer what's in namespace std::ranges; the rest are
      historical".

Cheers,

Chris