I share the sentiments of the responses above. Hardware has a feel that software doesn't. But the gap is closing with these software hardware hybrids hitting the market (the Beat Thang is an interesting standalone unit with a software complement).
I'm going to rant now.. you can ignore the rest of this post if you don't want to hear another lame forum user speculate on how AKAI can improve the situation.
If Akai wants to stay in the game they need to do two things:
1) Provide more transparency about their software updates by listing upcoming bug fixes/features on their website with approximate release dates. Why not start a beta forum or hold an online survey for registered users to vote on the next OS features? I'm sure AKAI could put its large user base to work for it by testing out beta OS stability and doing its customer research. Think of it as cheap labor like those Chinese workers you pay to assemble your machines

2) Provide a better way to communicate with software using USB... maybe some kind of streaming function to get your tracks into your chosen DAW that can sync with more timing accuracy than MIDI? Or how about a VST application that plays back MPC sequences and sample packs?
As a side note, AKAI could become a leader in the industry by venturing into uncharted territory. OSC protocol is stable and more efficient than MIDI and is already widely used in a number of software applications. It would be pretty amazing to apply the MPC's strength in sequencing to send/receive OSC messages that can be used in audio/video applications. It would open the MPC to a whole new generation of electronic musicians. ok I am dreaming
The apparent decline in hardware is due to the amount of choices out there but regardless of this flux in the market AKAI will still maintain a huge user base. If AKAI was smart they would try to capitalize on these older machines that users refuse to replace. How about re-purposing some of this hardware with a (free/paid) software update? Or releasing the code as open source? There are smart guys out there like JJ who can add features or create different applications for the unit and AKAI won't have to do a thing after the initial release. It would be easy to go open source and could restore some trust in AKAI with the community of electronic musicians. Best of all AKAI would not have to deal with complaints as they could just say "fix it yourself" which is better than their recent harsher tone of "deal with it!". At least there would be hope to get the machine closer to being bug free. AKAI, we already love the components and build quality so just give us control over our machines!
I was looking into the MPC 5000 but my biggest issue (as a longtime MPC 2000XL user) was always AKAI's lack of support. I love the work flow of the MPC but no matter how great of a programmer I was.. I still needed to get my sh*t into a DAW. Since I am already accustom to using other tools in the process, I'd rather find a solution that integrates well into my DAW. Sorry AKAI but it is the truth. Stand alone units can only go so far on their own in 2009. Let's find a better way of integrating the MPC into the PC world.
I'm personally waiting until NAMM to see what is going to happen before I make a purchase decision. There are a lot of things set to drop this January like the Linn Drum II, Beat Thang, Maschine OS update (that will include MPC program import), and hopefully something more from AKAI than a pimped out MPC 2500.
1-0ne