Might be my optimistic fan feeling, but despite what I used to hear a lot a few years ago, it looks like Goldsource modding is absolutely not dying, but alive and well instead.
I've for long thought that the only reason other people and I were sticking with this engine was the laziness to move on something new/nostalgia/die-hard fanaticism and that we were a slowly disappearing niche of irrational fans refusing to realize that moving to a more recent engine can only be better.
Now I think it's utterly wrong: to stick with Goldsource is a reasonable choice in its own right. The time tested graphical aspect and gameplay are enjoyed by many people, some of them just discovered the engine recently. Modding is far more simple than with some other recent engines.
The community is active: Sledge, Jackhammer, Xash, Cry of Fear, Paranoia 2 are relatively recent projects. Not to mention mods currently being developed.
And it isn't simply an active community of die-hard fans. Goldsource gets a fairly decent media coverage: Half-Rats has been featured in a British media, Sven Co-op got a large media coverage and an unprecedented number of players for its 5.0 release, Goldsource mods frequently get featured in ModDB and RTSL awards…
I'd say, if you`re not concerned about monetizing from your mod, go ahead and spend your time with GS, it's going to be here for a long time no matter what skeptics say. With a few graphic engine tweaks here and there, Volvo could probably still pull enough money out of it, but alas...
P.S.: Source/SDK is old AF too.
I like source for the amazing Input / Output system, and you have a lot more control over the AI. But it takes more effort to create something with the WOW factor
I enjoy some time with Unity for a small project. But it's not big and something that could be recreated on a custom engine.
Programming a mutator for KF2 (UE3) and making an unfinished Doom II MAP01 remake (not a copy/paste, job done, cya, but a complete overhaul to fit KF2's theme) also was fun.