I claimed not, as 'reduce' was already the "list to single value" generalisation. As evidence, and to pass the time on my train journey, here is a version in Python:
def fnreduce(sequence, initial): return reduce(lambda x, y: lambda _: y(x(_)), \ sequence, lambda _: _)(initial) fn_list = [ lambda a: a * 2, lambda a: a - 1, ] assert fnreduce(fn_list, 2) == 3
Replacing y(x(_)) with x(y(_)) would apply the list in reverse order, and thus fail the assertion. I just hope you like lambydas.