I agree with the sentiment about making the playing field as equal as possible. There is a line at which customization becomes cheating -- like maybe changing pickup sounds to tick down so you can hear when an item is respawning or making opponents visible through walls. Changing enemy color, particles, bolt reload sound, etc., however, in the grand scheme of things, I don't think have anything beyond a subtle impact on outcomes. Maybe two extremely high, equally skilled players, where a single mistake can sway the outcome of games, might see some differences depending on some very minor custom detail, but for the vast majority of players, it won't.
Also, as far as I'm aware, if you disable the IC/BR hum on your side, it has no impact on the other player. They still hear it whether or not you hear it. The hum is not designed to annoy the player who is holding IC or BR, it's to give information to the other player. You learn three things when a player is holding IC/BR out: 1) their general location, 2) that they have an IC or BR, 3) they've got IC or BR out as their current weapon. From these bits of information, you can infer other bits of information (like, if they just spawned, you might know what other items they may have which are located near the IC or BR pickups). These are all useful bits of information during a game. So if a person disables the hum on their side, that's sort of stupid ... Unless they're very cognizant of the hum they're emitting (even though they can't hear it), they could easily forget about it with it disabled.
All the customization features are seen as a positive by a lot of players. So there's no silver bullet here. You may satisfy some players but also upset other players. Personally, I'm on the side that thinks customization (to a certain degree) is a good thing. There's pleasure in customizing your game to exactly the way you like it. Just like anything else in life. The customization "issues" are minor, in my opinion, and probably do not impact how many new players Reflex is attracting and retaining.
I think the much more important point you raised is making the new player experience less brutal. AFPSs like Reflex, QuakeLive, UT, etc., have a brutal learning curve. This is the nature of AFPS and to adulterate that would be to lose what makes an AFPS game an AFPS.
The best way to address this issue is match new players with other new players and make sure they can play often and continuously (without waiting around). There's nothing better than learning a new game with people of similar skill level. Right now, the active community is probably more skilled than not and there's frankly not that many active players in general. So what few games are being played are sometimes between players of much different skill levels.The addition of matchmaking should help with this problem.
I think Reflex needs to achieve a "critical mass" of players at which point it will start to grow more rapidly. The goal should be that any player can start Reflex and, within 30 seconds, find a game with equally skilled players of their mode of choice. A few months ago, I did some estimating that given four modes (1v1, FFA, TDM, and CTF) and if we assume active player skillset is normally distributed, that in order for players to find a game within 30s via matchmaking, we would need something like 3,100 players playing at any given moment. Right now, Reflex is pretty far below that but my hope is that as the game moves out of alpha and into beta, the game will be much more polished, have more "identity" and "personality", and you start to see the community grow pretty rapidly.
Some secondary things I think that could help with attracting and retaining new players: 1) More resources for new players like training maps, videos, guides (although, there's already a lot of quality resources available -- maybe making them more prominent?), 2) more social features integrated into the game (like a chat lobby on the main menu screen?), 3) easy ways to share in-game moments with friends (meaning, easier to use replay editor).