well, i just discovered it about a week ago, so I hope my opinion sounds honest to you.
In terms of commercial games- yes you can export to html5, so at least in theory you can make commercial games that run on anything.
Yes it is very similar in some ways to construct2. But it lacks some basic features that keep it a lower class engine in comparison.
At the same time- it has some interesting native support features and a native linux port that set it a part. The completely free license with no limitations is the other thing that sets it apart. The half open sourced nature is another thing that sets it apart- developers can pick up an SDK and start writing their own plugins without paying a cent to anyone.
Having said that, I would also have to point out where this engine has weaknesses at the moment.
Both construct2 and mmf2 have a built in sprite editor - with painting tools. This engine does not have any painting tools - so you are forced to always import graphics before starting coding- no way to quickly make some placeholder sprite.
The community of both other engines is much larger- with a lot of tutorials covering everything in them in english- both in reading and video form. So both of the other engines are better documented, much more tested and widely adopted. The reason for that is that those guys have way more resources and 4ian is a one man army here.
Now having said all that, we must remember that this engine has a lot of potential to become something very special.
First of all the other two engines are tied to windows libraries, thus their developers will likely never port them to linux and mac.
In the long term that gives gamedevelop a chance to stand out- as it is already running on linux and that means it can probably be ported to mac as well. This is a market share that gamedevelop can capture before construct2 gets to it.
After the disaster that windows 8 was, a lot of people are staying on 7 or switching to mac/ubuntu. There are exciting developments with Valve releasing their own operating system. If gamedevelop manages to get more developer resources, at least one or two commercial games made by it, and more users know it well enough to make tutorials for it (not just 4ian)- with tutorials actually being in english. Then I see it get widely adopted.
Right now it is lagging behind and it must keep above the water. Keep evolving.
There have been other engines in the past that were similar to construct2 and mmf2- like “darkflow” for example. They all died out. 
I personally want to see this engine become widely used and really competitive with the other two.
The reason is that competition is a good thing for the users. It forces the other engines to compete as well.
Now if this engine ever becomes open sourced- well I think it will probably get a lot of developers to pick it up and users too. But it still needs to sell something in order to make money. Without the funds, we can not buy engine devs time to work on it - and that keeps it a hobby project for them.
History shows that hobby projects grow in size really fast when they become the hobby of many individuals 