If an AV's BB can block any ransomware I consider it a success. BBs don't have a lot of time to figure out what that programs up to until it's to late.
Actually, a Behavior Blocker has plenty of time to calculate different scenarios based on the behavior pattern of the sample since they work through interception of API calls (which is then logged) however they can update the score systems and perform comparisons before the function has been redirected to be completed (since the behavior blocker code becomes executed before the function call passes through to kernel-mode by NTAPI wrapper).
If they don't detect it's not because they "don't have a lot of time" (suggesting they would have caught the sample after awhile longer), it's because it won't detect it since it doesn't link the behavior pattern to representing a file with suspicious/malicious integrity, therefore it'd be a clean miss.
Regarding ransomware, not all work the same and new improvements are being added all the time - since some samples use static linking as opposed to dynamic for whatever encryption library is being used, it also makes it much more difficult to monitor. Therefore, most anti-ransomware which are well-developed and truly evolving around dynamic analysis will just monitor file modification (intercept on file write attempts - can be done with a device driver or via hooking of critical NTDLL functions) and then identify when a file is being encrypted (if it can), and the such. However some ransomware is more advanced than others (for example Petya will infect the Master Boot Record) meaning depending on the functionality of the sample and depending on the monitoring capabilities the component has in that particular software will depend on whether the zero-day ransomware will be blocked.
For example, some ransomware may be blocked by Emsisoft but might be missed by Kaspersky Anti-Ransomware, whereas some samples may be detected by Kaspersky Anti-Ransomware and missed by Malwarebytes Anti-Ransomware.
Most or all. As Emsisoft did in that video.
That's because they probably use a real dynamic method like the one I suggested above, as relying on only generic signatures for ransomware identification (and some anti-ransomware do just this and nothing else)... And monitoring file modification to identify suspicious modifications from an unknown/untrusted program is very clever and reliable if done correctly.