That being the case, I retract my statements regarding pixel-precision as an argument against the Spearer, in the same way that I still don't believe that the Grenadier is necessarily "backroute-prone", and to give marks against it in this regard seems unfair. How is it any more backroute-prone than any other skill?
Some skills are absolutely more backroute-prone than others, and this is not really debatable - it's well established over years of level designing. Walker, Builder and Stoner (and in L3D's case, Turner) are the worst offenders.
Yes - you can often modify a level to make it backroute resistant; this does not change that some skills lend themselves more to causing backroutes than others do. Yes, basically every skill
can lead to backroutes, but think about how many backroutes revolve around eg. a Floater or a Swimmer or even a Digger; compared to how many revolve around the higher-risk skills mentioned before.
What
is still debatable is where Grenaders (and the other skills) fall on this spectrum, and how mitigatable their risks are. And indeed, modifications to the skill's physics to reduce that risk
are an option if the current iteration is deemed to be too broken - it is not a matter of "the implementation as it stands right now must meet expectations".
One thing I will make clear is that I do consider this to be an important point. NeoLemmix's main focus is on puzzles; a skill that does not lend itself well to this, no matter how fun it is and/or how well it lends itself to open-ended exploratory and/or execution difficulty levels, is not a suitable fit for NeoLemmix. Going into this, the Grenader was my favorite of the skills, so this is not that I am "biased" against it in any way - it's that I genuine have concerns about this point with regards to it. Assertions that "this doesn't matter" or even more so that "all skills are backroute prone" (which ignores the difference in
extent of how backroute prone some are compared to others) are ultimately going to be meaningless - if the evidence points to that it
is a high risk skill and little is offered in the way of reasonable mitigations that don't give away the solution; then ultimately the skill is very likely to be rejected no matter how many assertions have been made that "this doesn't matter".
But again, let me be very clear - I am not at this stage saying "the Grenader is too broken and must be rejected". All I am saying is that this line of questioning needs to be explored further - and that while the Grenader is my biggest concern, it should be explored for all four skills to be safe.