I think those trying to prohibit bad methodology easily fall into talking about it as being by its nature always useless. that people naturally do that doesn't damn their position entirely.
I think it would help to make explicit your position (others too but first yours) tell me how much I have right....
1) you think that there can be philosophical progress (i.e. not the strong position being argued against above)
2) you think that progress tends not to happen in the field of philosophy (I presume because of how philosophy forms free floating ideas rather than ones 'grounded' by their attachment to empirical evidence)
3) you evaluate 'progress' and 'useful' in a different way to Richard and HA, Richard would find a interesting logical debate 'useful' you would ask if the logic can be used to make a car or feed the hungry or whatever. to get slightly more philosophical - maybe it is 'empirically verifiable truth'?
4) To try to use data created by a feild you dont trust to create good data to prove that the feild doesn't create good data seems unlikely to be unproductive (shutting the door to the 'why don't you argue from the literature').
A non reductionist might talk about X where X is specifically defined as 'stuff that cannot be reduced'. The reductionist hears the term X and starts to argue how it can be reduced. Point is that the non-reductionist is fundamentally talking about something different.
the solution is that one can use a on reductionist framework to consider the issue (with an associated shift in definitions of words etc) and one can use a reductionist framework or one can take a position somwhere in the middle. In my opinion the reductionist one is more useful - but if the objective is to "understand the non reducible nature of the soul" (trying to be extreme) or anything along those lines then I guess non reduction might be a way to go.
Similarly, to a non-reductionist a world which is identical but with no qualia is an interesting topic and is conceivable - to a reductionist it may be 'inconceivable'. Just a different definition of the fairly flexible thing 'conceivability'.
mtraven creates a good example,
it would seem mt doesn't seem to care about reductionism. Effectively saying it achieves nothing. On the other side a reductionist would wonder what he was going to achieve via analysis of mind without any sort of reduction.
Similarly we have a semantic debate about if hands without fingers are possible - well to some people it is and some it isn't depending on how they define hands - and whether that should or should not be the case is a pragmatic matter.
So in a fairly pure form you could have a person who denies that 2 is reducible to 1+1 and that 2 can exist without 1 - because he understands the concept in that way. Another person may say thats 2 is only 1+1 and that there is nothing theoretically interesting in the change entailed by having (1+1) appear everywhere that 2 is now.