Staying on Windows 10 1803 is really not recommended at this stage as that branch went end of life a year ago, so no longer gets security updates - see the Microsoft Windows 10 Home and Pro lifecycle web page.
As all Windows 10 users are entitled to upgrade to newer feature branches free of charge, developers are fairly aggressive in ceasing support for older branches to minimise their test cases and support burden. Some bugs and mis-features simply are not fixed in older branches. Whilst it is true that some Windows 10 feature branches had serious issues at launch, I believe the best way ahead is to let others take the early adopter pain and upgrade after 2-3 months via Windows Update, which ensures you are not upgrading against a compatibility hold that applies to your machine.
You are right that 2003 is not a version - it is 2004. Microsoft has moved to using ‘yyHh’ designators from the latest feature branch onwards, which is 20H2 (second half of 2020) because the month part of the ‘yymm’ designators became rather arbitrary when release dates slipped.
At this stage, all my Windows 10 systems and virtual machines are on 2004. 20H2 is a relatively minor upgrade over 2004, but I’d rather let companies like Steinberg and Adobe test their software against it rather than upgrading and hitting issues.
Nobody is going to tell you that your old and unsupported Windows 10 is OK to run Cubase Elements 11 now, let alone future minor releases. You can test it yourself once Steinberg offer trial licences in a few weeks’ time. However, I’d urge you to get back onto a branch with widespread vendor support by backing up your system, upgrading the drivers for your key hardware (especially graphics cards, audio devices and any Wi-Fi and Bluetooth adapters, which are the most common devices causing compatibility issues on upgrades), then upgrading to Windows 10 2004. This puts you on a branch that is supported for another 13 months, with a relatively simple pathway forwards to the latest 20H2 branch once that branch is more mature.