Choice-based games are orders of magnitude more popular than parser games, to the point where some of them are commercially viable, mainly the Choicescript/“Choice of Games” ones.
The IF competition is the most important competition for the community, and probably the best way to get feedback on a game. Although, if you are doing a popular Choicescript game or a graphically enhanced game which stands a chance on other platforms like Itch etc., you might not need the IF Comp for exposure.
There is a relatively large overlap between authors and players.
Several people feel that a small number of thoughtful, in-depth reviews is worth more than a medium or even large amount of commentless downloads – although there’s presumably some threshold where that switches around. 
It is an activity that should mostly be its own reward, with critical acclaim and a positive reception as a bonus in the ideal case.
You can get a rough idea of the dimensions by looking at the number of ratings on past jams and comps.
ECTOCOMP 2018: 22 entries, 212 ratings (across all entries, so on average 9.6 ratings per game)
ECTOCOMP 2019: 21 entries, 188 ratings
ECTOCOMP 2020: 17 entries, 293 ratings
ECTOCOMP 2021: 23 entries, 310 ratings
ECTOCOMP 2022: 35 entries, 678 ratings
ParserComp 2021: 18 entries, 355 ratings
ParserComp 2022: 18 entries, 308 ratings
The quantity of votes cast is also noted on past IF Comp results pages for each game. For example, in 2022, the numbers ranged from 20 to 122. Of course, there are also people who play but don’t vote.
Here are some related threads:
-
IF economics
(from 2020) -
IF User Base
(2013)