Thanks for your feedback!

I am not sure if it is really a disadvantage to be stuck with the Lion until the end-game. Preventing the game from reverting to regular FIDE sort of forces this on us. I must admit that I don't know what consequences this has for the drawishness of end-games. It would indeed be bad if KLQKL would be a general draw. I guess I should investigate this by building some tablebases. (5 men on 8x8 is quite easy, and even 6 men might be doable on modern hardware, but my tablebase generator should be modified to take account of the Lion-capture rules, and the possibility to do hit-and-run captures.)

If KLQKL is generally won, then KLPKL might be generally won. And if small advantage can still be converted to wins even in the presence of Lions, there isn't any harm in the Lions always being there.

Your suggestions for wider boards are also interesting. I picked 8x8 so that it would be easy for people to play this over the board using standard equipment. (Dressing up one Knight with a rubber band, or putting it on a pedestal to identify it as Lion. Or just use any object of which there is a pair at hand, like a stack of Chekers chips or pepper and salt shakers in case both Knights are kept.) 

If the board were expanded to 10x8 I would rather add Shogi-like pieces such as Dragon (RF) or Horse (BW) than Capablanca pieces. E.g. put Lion next to King and Dragon next to Queen. Chess players would have to learn two new pieces rather than one, however, which is a bit at odds with the goal to lower the threshold. Even though I agree that it would likely give a much better game, both through the wider board making the short-range Lion less dominant, as well as offer more interesting trading probabilities for Lion vs other material (in particular Q+D).

My reason for not allowing promotion to L was to avoid complicating the Lion-capture rules by the possibility for there to be multiple Lions.

In any case, it seems essential to make some end-game tablebases, to check if this game would need tweaking of the Lion-capture rules to prevent it from being drawish.

Btw, what do you think about my defining of the protection of the Lion in terms of pseudo-legal recapture? In Chu Shogi this is natural, as exposing your King to capture is not illegal there, just stupid. But in Chess it would be illegal. So it is debatable whether a piece that can be recaptured only by a pinned piece is protected or not. Even in the FIDE game this gives rise to a subtlety: it would be protected against capture by a King, but not against capture by any other piece. For simplicity as well as conformance to Chu Shogi I treated the Lion the same as a King here: you cannot make distant LxL capture if a King could not have legally captured that Lion.