I moved directly from DBF to MS SQL Server, many years ago, and started with version 2005 SQL server - and have moved through many newer versions since. All were fine, and worked OK!
I have never had any issues at all, and there are also free versions to suit a developer and a small local network, as well as a small single machine.
Currently I run my son's large engineering database in the cloud with Azure - it was easy to upload the one million plus records with a free tool MS supplied. This DB started life in a Vulcan app and then was moved to LINQ with C# and "Linq 2 SQL".
Now when I open VS 2015 / 2017 this cloud DB is easily available at the click of a mouse. Quite brilliant. See attached images :-
To make such a DB available to any / many user(s) - and the very latest MSSS version we don't need to setup a virtual machine - just opt for a SQL Azure database. I get mine (Azure time) free as part of my Prof VS package. I also get all the SQL Server versions too.
When I return from Cologne (going in the morning) and get back to writing and researching the X# new eNotes eBook, I will do a full explanation of what is on offer from Microsoft as regards SQL. They do a local simplified version a bit like SQLite.
Personally, I feel that cost is not a good reason to keep clear of MSSS, as in recent years MS have done some good things to open up to small users the power of their Big Brother (and very mature) product.
I have tried SQLite and it works, but I only used it because at the time there was not an alternative for Mobile and WinRT devices. I definitely prefer the full MSSS offering.
Oh! - just thought, I would always use a flavour of SQL which will work through Entity Framework 6 or later. EF6 is just too good to miss just because the flavour of the SQL server does not play ball with it. To me this is a BIG concern and issue.
SQL is very different from DBF and ADS, it is a completly different approach.
It is now more than 10 years I'm using SQL in different flavours, but my main database has remained DBF.
I have selected SQLite only for local and configuration data because it is fast and easy to copy.
My decision for PostgreSQL has been because it seems to be the most powerful of the free databases. With MySQL I had several data corruption issues because of power failures, but they occur on my customer sites (small business).
The most stable product I know is Oracle, but it haves a cost, and it is not so easy to configure as I would like.
From Harbour, there also exist LetoDB as an opensource alternative for ADS. There's also a fork of the former as LetoDBf and some ADS users are testing it as drop-in replacement without changes from their existing code with far better performance than their current ADS implementation as they have reported. I'm also using it (LetoDBf) in production and very satisfied with the stability and performance but I don't have experience with ADS.
Currently I run my son's large engineering database in the cloud with Azure
Apart from my built in resistance to almost anything Microsoft offers, which for me is reason enough to stay away from Azure or MSSS, I have one other objective. The great thing of DBF files is that it is so easy to just work on one dbf (e.g. replace it with a backup file) and -more important- it is so easy to copy it to other PC's. All off line. When I travel everything I may need to help customers, my employees or just to do some work, is on my tablet-laptop. And in most places I travel, internet is absent or dreadfully slow (I have 150/15 here). So a Azure based database may be a good idea if you don't have a server (and/or have a fast and very reliable connection) but it's not portable.
I acknowledge that having data via cloud services has it advantages. But as soon as it's the only option (e.g. Office 2016 help, this forum instead of the NG, Azure databases) I don't want to use to use it - or better said: I can't use it as quite a few weeks in the year we will not be able to!