I think that is a wild assumption, as in this stage people don’t really use it for payments yet. Most transactions will just be from exchange to personal wallet, and I don’t expect any significant increase in transactions because of the lower fee, since exchange fees generally already outweigh xem’s inherent fees. For the higher demand of XEM due to the lower fees I would use the same argument: fees are not a big issue yet and I think demand is not constraint by it at the moment. In fact, demand might even go lower because there is less to harvest (judging from questions on this forum a lot of people got into it because of the harvesting).
Even so, to me the reduced fees also do not have a purely negative effect. It is just that I can imagine that not everyone agrees with that, yet currently it seems that whatever developers want to change to the system will be changed, without the need of the backing of users (harvesters) in a formal way. If that will change by means of a voting system bound to importance rate, that would be great, but as I understood from mizunashi that is still a plan for the future.
In the meantime, it seems very uncertain to invest in xem as users cannot control or are uninformed about exactly what features will be implemented. And fees are a very important part of the system, whether the change turns out to be positive or negative. If that remains the way it is, there might be forking hazards if changes are implemented that users are unhappy with.
For example, imagine the extreme case that developers plan to double the amount of xem in existence, with the larger part reserved to themselves. I trust they wouldn’t, but in the theoretical case they could just implement that (even within the still hidden Catapult release), and all users would just have to deal with that or use a fork of their own. That doesn’t seem healthy to me.