No. It's impossible to reduce the latency of transmitting the communication signal via fibre between the server and the client (which makes up a significant portion of the latency on WANs). This is compounded by routing and switching overheads and is exacerbated by congestion etc.

 

Yes.  It's been long since established that 100 ms is the upper limit on human perception that a visual response is instantaneous (see Miller, R. B. (1968). Response time in man-computer conversational transactions. Proc. AFIPS Fall Joint Computer Conference Vol. 33, 267-277).  Given that most links out of the country to US/UK datacentres are measured to be at least 150ms, a CIO will usually pay more for the increased productivity that is achievable with a lower latency link.  Moreover, there are many other good reasons for keeping things local e.g. higher throughput, in-country physical presence of the data, different legal jurisdiction.

Your message highlights an interesting point though: The importance of education in the market - people don't understand the concept of latency and so think it's something mitigatable when it actually is not, especially for interactive applications.