SQL Server - Convert date field to UTC

With SQL Server 2016, there is now built-in support for time zones with the AT TIME ZONE statement. You can chain these to do conversions:

SELECT YourOriginalDateTime AT TIME ZONE 'Pacific Standard Time' AT TIME ZONE 'UTC'

Or, this would work as well:

SELECT SWITCHOFFSET(YourOriginalDateTime AT TIME ZONE 'Pacific Standard Time', '+00:00')

Either of these will interpret the input in Pacific time, properly account for whether or not DST is in effect, and then convert to UTC. The result will be a datetimeoffset with a zero offset.

More examples in the CTP announcement.

I do not believe the above code will work. The reason is that it depends upon the difference between the current date in local and UTC times. For example, here in California we are now in PDT (Pacific Daylight Time); the difference between this time and UTC is 7 hours. The code provided will, if run now, add 7 hours to every date which is desired to be converted. But if a historical stored date, or a date in the future, is converted, and that date is not during daylight savings time, it will still add 7, when the correct offset is 8. Bottom line: you cannot convert date/times properly between time zones (including UTC, which does not obey daylight savings time) by only looking at the current date. You must consider the date itself that you are converting, as to whether daylight time was in force on that date. Furthermore, the dates at which daylight and standard times change themselves have changed (George Bush changed the dates during his administration for the USA!). In other words, any solution which even references getdate() or getutcdate() does not work. It must parse the actual date to be converted.

If they're all local to you, then here's the offset:

SELECT GETDATE() AS CurrentTime, GETUTCDATE() AS UTCTime

and you should be able to update all the data using:

UPDATE SomeTable
   SET DateTimeStamp = DATEADD(hh, DATEDIFF(hh, GETDATE(), GETUTCDATE()), DateTimeStamp)

Would that work, or am I missing another angle of this problem?