Chess tactics too easy and straightforward

As a veteran of chess.com, I see that they're puzzles offer much more than just what lichess tactics offer. As a result, they are much more difficult.

Why does lichess tactics always involve a move that checks the king first? Oddly, at least 4/5 tactics involve just scoping out a move that checks the king, which leads to a predictable string of moves. Chess.com offers the catergory of attack as well (when you finish the tactic) such as "fork" or "xray" or "perpetual check".

I have not seen a tactic yet that tries to put a lose into a tie, or a move that actually just increases your overall position on the board. Its always checkmates.

If the software you use to create these tactics can not detect moves that involve positions other than "attack the king" than I'm afraid it becomes a very boring and harmful deterrent to actual chess training.

TL;DR Tactics need to be revamped

You are somewhat right, quite often the tactics is just hidden checkmate, not always though. For example, I have just come across this one: http://en.lichess.org/training/9423

We're working on it.

The hardest part is developing an algorithm to detect these kinds of tactics. A good example of a tactic that you could expect in the high levels are the philidor and lucena of a rook endgame, which are not trivial in any sense.

What we will probably eventually see is the addition of "classical" tactics, which are tactics that have been around for 100s of years and have taught masters for ages like this one here http://www.thechessworld.com/learn-chess/33-problems/397-7-most-famous-chess-combinations

I personally think that they should add an endgame database to their tactics set to train for endgame. I know nalimov is commercially used but I believe that syzygy is superior for some reason (I dont do chess programming so I would not know)