From a purely logical standpoint, if what the official in this video is saying is correct (i.e. after the end of 2016 the only way to immigrate using an economic pathway is through EE) then there is no way that the required CRS score cannot come down, although based on my calculations I cannot see it decreasing hugely. If there is something wrong with my calculations below, I hope that a senior member of this forum can show me the error of my ways!
Let's take 2015 as a starting point, as we have good data for the whole year provided by CIC. The Canadian Government's immigration target for 2015 was between 260,000 - 285,000. Considering only the classes that it is possible to apply for through Express Entry (FSW, FST and CEC) the target totals were 68,000 - 74,000. I have not included Provincial Nominees in this figure as, although there was a minumum target of 46,000 provincial nominees, only 3960 ITAs were issued to provincial nominees through EE during the whole of 2015 (approx 152 per draw) which I'd argue has a negligible effect on the average required CRS score. Many PNP nominees seem to go through a different process, not through EE (I have little knowledge on this, perhaps someone else can help and tell me if PNP is changing and more provincial nominees will be starting to go through EE in the future). Also remember that there are numerous programs (Federal Economic Caregivers, Federal Economic Business, Quebec Skilled Worker etc) which take away from the economic migrant category that are not applied for through EE and cannot therefore affect required CRS score or numbers of ITAs given.
There were 31063 invitations sent out through EE in 2015. Working on the (worst case scenario) lower figure, you can see that EE invitations were 36,937 short. I can only assume that these came from the old system prior to EE coming in. If everyone had been selected from the Express Entry pool in 2015 instead of through the old system, there would have needed to be an average of 2615 ITAs issued per draw, whereas in actual fact there were only an average of 1194 ITAs per draw. Let's look then at the first draw of 2016:
On January 6, 2016 there were 1,463 ITAs issued with a required CRS of 461 points. Had there been no pre-Express Entry candidates taking up the economic quota, there would have been (on average) 2615 ITAs issued instead. That would have put the required CRS total down to somewhere between 440-449 (figures aren't available which break this down further to give an exact score).
The 2016 figures show for a lower proportion of economic migrants than 2015, despite an overall higher target. This would appear to be as a result of a large increase in refugees being taken into Canada. The targets for FSW, CEC and FST migrants for 2016 is between 54,000 - 59,000, giving an average target ITA of 2076 - 2269, based on 100% of FSW, CEC and FST migrants going through EE. As CIC is, however, still working through the backlog and in 2015 only 54% of the economic migrant quota was taken up with EE profiles, it is entirely unsurprising to me that the average number of ITAs has been as low as they have - currently 19 draws, total ITA of 19,292, with an average number of ITAs per draw of 1015 (based on 2015 percentage of applicants EE vs old stream (54%) I'd expect somewhere between 1121 - 1225 ITAs at this stage, so we're not far off what I'd expect).
In a hypothetical scenario where the backlog was cleared tomorrow, I'd expect there, based on 2015 figures, to be around 9,868 and 12,568 invitations available before the end of 2016, which would be issued through Express Entry. There are 8 draws left this year so I'd expect ITAs to average somewhere between 1233 and 1571 between now and Christmas. Obviously the longer it takes them to clear the backlog, the lower the average number of ITAs will be, the corollary of which of course is that the required CRS score will be higher. If they've proceessed fewer non EE applications than I expect they have then CRS will possibly go higher, perhaps even towards 2000 per draw.
The best prediction that we can reasonably make for 2017 at this moment is that the number of economic migrants will generally stay consistent with 2016, as we have no further information telling us that they intend on increasing this. Please let me know if I've missed any important news on Canada planning to increase migration on what was already a record 2016! If we then assume between 54,000 - 59,000 EE ITAs to be issued that's 2076 - 2269 ITAs per draw, requiring a CRS score of around 450 - again not far off the 2000-2500 which was suggested by the CIC employee in this video. The unknown factors, of course, are the coming changes to the CRS system, such as Canadian students being given more points (will put upwards pressure on required CRS score and PNP/job offer points coming down, putting negative pressure on CRS requirement) as well as the actual immigration targets for 2017.
In summary, even in a best case scenario where we have over 2300 ITAs per draw from October and throughout 2017, I cannot see CRS dropping significantly or at all below 440. Perhaps I am misunderstanding the numbers (and if I am, someone please correct me), but 5000 ITAs being issued per draw with CRS lower than 400 is a pipe dream at this moment. Even if all PNP nominees went through EE, yes the number of ITAs would go up, but also the required CRS will rise accordingly as those people will all get nomination points - leaving little to no net effect on CRS score requirements for non PNP candidates. The problem that people in the 400-450 range have is that there is such a large number of people in that band (more than 13,000 as of January 2016) even if you quadrupled the number of ITAs that are being suggested by the employee in this video and based on my most optimistic calculations which tend to agree with the CIC employee, you'd still not get below 400 CRS.
We're looking, I believe at a situation over the next year or so where there will be over 2000 ITAs being given per draw and CRS being around 450. This will then be affected by the forthcoming changes to CRS score which I believe will push the requirement up slightly, in the broad region of 460-480 in the long term. That's my prediction - would be interested to hear thoughts from people on this.