Simply improving a processor isn't going to change the fact that what people want are low cost processors without vendor lock in.
This is what they say. Then they put Apple on desks and in bags, and buy iPhones, not because it's lower cost but because it's viewed like a perk (like free lunches, snacks, or an on-site gym, etc) - and it's relatively cheap compared to payroll costs in general. What they say is largely irrelevant because they don't know what they want, and in the absence of wanting something else a purchaser will say they need lower cost. When a purchaser says they're cost sensitive it really says they don't know what they need and aren't in the business of brainstorming. This doesn't mean there's no opportunity for innovation outside of cost-performance. Sun benefited from its stackable pizza boxes; people greatly preferred having a stack of those in a corner to the other common alternatives at the time, even though they cost more. They also ran SunOS. Sun wasn't popular with purchasing departments either, but they had no choice because if they bought Pentium PCs or IBM AIX based machines their technical staff would go work someplace else. Sparc was only briefly, if every, best bang for buck in terms of performance. The problem with Sun was their product innovation completely stopped and all they did was tweak the underlying technology. Oracle is just as bad if not worse. This means they have their market locked in, but it's an ever shrinking market that can't be grown.
Imagine instead that you have a rack where you can add and remove boards as you see fit. Need 4 more CPUs each with 16-64 execution units, insert a board. Need another 64GB memory? Insert a board. Need a pile of ethernet ports perhaps with a builtin soft configurable switch, add a board. Or FC-AL, or anything else. And, more interestingly, want to remove a board? Push the stop button on it and the kernel begins releasing all resources on it used - unconfigure interfaces, power down CPUs, or deallocate (possibly page out) memory. When it's safe to remove a light comes on and you pull it out. A power supply might take up four slots; need redundancy? Put in 2 and set the mode to active slave. Or stick in a video board and run it like a big fat desktop. Such a system might cost more than the equivalent x86 servers, perform less, give less bang for buck, but it would have reasonable incremental expansion costs and flexibility. It would find an immediate customer base in academia and technical markets - the pizza box of the 2010s.
In the case of the Sparc T5 they're once again just going after cost. Nothing wrong with that, but more telling is there's no product innovation going on.