After all, just using our eyes the fact that the sun circles the earth is perfectly valid. It takes observations, with some kind of vision enhancement (telescopes) of a number of different celestial bodies in the solar system to gather the data that refutes a geocentric view. Especially when that view is a mature and accurate theory. For trivia's sake, if you go to a planetarium and enjoy the show about the solar system, the entire setup uses a geocentric model for the same reason that geocentric theory was developed. It is the best way of dealing with bare eye observations of the solar system from the earth, especially with modern computers. Heliocentric theory simplifies the math, so it fits Occam's Razor.
The worst thing is that laying out a case like this, even on a simple matter, to overcome existing beliefs, becomes a wall of text even handled relatively briefly. In RL, this took me minutes to figure out once I started studying the relevant rules sections. So, long post ahead.
The first thing we will look at is the rules on Declaring Assaults on page 33. Looking at the summary, the second bullet of 1 is "Declare which enemy unit it is going to assault". The right column of page 33 has the rules for declaring assaults. The first thing is that all assaults is to measure the assault range. The next is the subsection on "Disallowed Assaults". There we find five bullet points and a conditional clause ("must assault unit you shot at" - will only use "must assault" for this from now on). Interpreting this, the first question is do we assume that GW used standard English and style rules writing this section. If yes, then we have five "hard" conditions that always apply and a "soft" condition that only applies in some cases.
Note: When you make assumptions to prove a theory, you also have to understand the implications of that assumption. Therefore, assuming that GW messed up the writing here and "really meant to have six 'hard' conditions" implies that anything written in the rulebook can be tossed out (RAW) in favor of interpretation (RAI) on the basis of "it should read this way, GW just wrote it wrong". That is a can of worms that tosses all of the rules away at whim.
So, the best reading is that there are five 'hard' (bullet) conditions and one 'soft' condition about only being able to assault the unit you shot at. The next assumption (these could also be called "interpretation") comes when deciding if a destroyed unit still counts as a unit for the purposes of "must assault". Of course, if the target unit is still functional on the board (IOW, doesn't count as a KP), then the "must assault" condition is triggered. So, we are left with two choices: A unit that is destroyed (which includes wrecked vehicles even though the model may still be on the board as terrain) either is still a unit in the next phase OR it isn't a unit, it is just a KP because it is destroyed. Although I think that making the assumption that units still "exist" in later phases is a bit iffy (it is a bit hard to declare an assault on something that isn't there, but that is what assuming the dead unit "exists" requires). Another issue is that even though "must assault" logically must be a soft condition, if you consider the unit "exists" even when it doesn't, this assumption shoves "must assault" back into the hard category because if you shoot, it doesn't matter what happens to the unit you shot at, you "must assault" it. Well, duh, this turns it back into the same problem as considering it a hard condition from the beginning with the same implications for the entire rulebook.
The next strand in the theory is the use of the note (Note) in the second column of page 67 at the end of the Effects Of Damage Results On Passengers section. It reads as follows: Note: remember that all models in a single unit fire simultaneously, so a sqquad cannot take out a transport with its lascannon and then mow down the occupants with their bolters. However, if a transport is destroyed (either result) by a ranged attack, the unit that shot it may assault the now disembarked passengers, if it is allowed to assault according to the assault rules.
The common interpretation considers this to be an exception to a hard "must assault" assumption, either from directly or indirectly interpreting the RAW conditional statement as a bullet point or hard condition. This can cause some problems reading the Note as written. Again, there are two cases to consider:
- This is an exception because the "must assault" is defined as a hard condition.
- This is simply a clarifying note because "must assault" is a soft condition.
Next, we see that a unit that shoots and destroys a transport may assault the passengers, "if it is allowed to according to the assault rules." Of course, this immediately counts as an exception to the hard "must assault" condition, but are there any other exceptions? Well, obviously the shooter must be in assault range (and have a WS), so this applies to the Disallowed Assaults section and its hard conditions. Why not just the "must assault" condition? Because the Note does not give any other trigger as an exception except the transport be destroyed in the shooting phase and the shooter shot at it. Two choices again, either all of the hard conditions still apply or none of the apply if the exception's trigger is met. Logically the first three bullets (locked in CC, ran, GtG) cannot apply because the unit could not have shot. If the other three all apply, then the shooter cannot assault.
However, if this is an exception, then a unit that fired RF or Heavy weapons and a unit that is falling back also must get the exception if they qualify by having shot at the transport and are within assault range of the passengers. So, to make this fit in the common interpretation, two more assumption have been made; the final phrase of the Note doesn't apply to the hard conditions and that it only applies to the "must assault:" hard condition.
For point 2 above, the Note is read as a clarification. This fits with the "remember" reference for shooting and then ties into the "However,..." for a relevant clarification since we are discussing transport vehicles. If "must assault" is soft and the dead no longer exist for game purposes, it is fairly easy to point out for non-vehicles because they are pulled off the board as the models die. For vehicles, the rules state that they remain on the board as terrain when wrecked (page 62). I've seen more than enough players who would point at the wreck and say that the model is still there so the unit is "still there", but you cannot assault it because it is wrecked! More seriously, it clarifies that even though the passenger unit was off the board at the beginning of the turn (and therefore not a target for shooting or assault normally), by coming on the board during the shooting phase, it definitely may be assaulted, subject, of course, to the assault rules, including being in assault range and using the applicable hard conditions. No need for extra assumptions, it reads exactly as it is written.
So, both positions must assume GW wrote what they meant or didn't write what they meant. Then they must assume that a destroyed unit exists in later phases or that a destroyed unit ceases to exist for game purposes except as a kill point. Finally, the supposed exception of the Note (which is cited by the defenders of "must assault" as a hard condition) requires two assumptions (does not apply to all hard conditions and only applies to a hard "must allow.
So, four assumptions to two assumptions (I'd say five or six to two, but not now), which gives the "must assault" as a soft condition that doesn't apply to destroyed units the benefit of Occam's Razor for simplicity. Notice that this does NOT deal with seriously judging the validity of some of those assumptions, but I noted some issues in passing above. Heh, in some cases the trump card in defense of a hard "must assault" was quoting the 4th Edition rules on assaults verbatim. For 5th Edition