Is this expert report wrong about basic kinematics? This question is about the application of kinematics in an expert report for Eirikson v. Breton, ABQB 2000 798 (archived), a judgement for a case where a woman was driving a car that was accelerating as it crashed into a parked fire truck, resulting in her death (para. 65). Her family sued the city, and the city was somehow found to be 100% liable for the incident (para. 77).
There are multiple expert reports, but this is the one in question:

[48] Dr. Navin is a engineer and director of Hamilton-Finn Road Safety Consultants Ltd. Dr. Finn provided an opinion with respect to the visibility available to Ms. Eirikson and the amount of time needed to react to the fire truck. He made a site visit in March, 2000. He concluded that the flashing beacons of the fire truck would first be visible about 610 metres prior to the conclusion. The fire truck would have remained in the driver’s view for the entire duration until collision, if the driver continued in the median lane. The right side of the fire truck would have come into the driver’s field of view at about 183 metres. Finally, the fact that the fire truck occupied part of the median lane would be apparent to a driver at about 163 metres. Based upon those assumptions, the time remaining to collision would be 6.5 seconds at 90 kph, 5.9 seconds at 100 kph and 5.3 seconds at 110 kph. The stopping distance needed is 3.1 seconds at 90 kph, 3.4 seconds at 100 kph and 3.7 at 110 kph. That would give a reaction time of 3.4, 2.5 or 1.6 seconds respectively.
[49] Based on the estimated volume of traffic, there would have been little difficulty in finding a suitable gap in the centre lane into which she could have moved given the amount of traffic. He arrives at this conclusion simply on the basis of a per hour traffic volume of 2,300 to 2,400 vehicles.

The bolded part seems to be incorrect. They divide the distance 163 m by each of the three initial speeds (90, 100, 110) km/h to obtain (6.5, 5.9, 5.3) s as the "time remaining to collision". They then calculate stopping times by dividing each initial speed by an assumed deceleration of about 29.4 km/h/s to get (3.1, 3.4, 3.7) s. They subtract the times to calculate the available reaction times (3.4, 2.5, 1.6) s.
Because the car is decelerating, the deceleration would increase the "time remaining to collision" and invalidate the first calculation. Instead, it seems like the analysis should be done in this way:
For an initial speed $v$, the distance covered while stopping by decelerating at rate $a$ is $\frac{v^2}{2a}$.
If there is distance $d$ available, it is possible to drive at the initial speed for $t$ time before decelerating if
$$vt + \frac{v^2}{2a} \le d$$
$$t \le \frac{d}v - \frac{v}{2a}$$
If we crunch the numbers, we get the maximum possible $t$ values as (4.989, 4.167, 3.464) s. Is this analysis correct?
Also, 29.4 km/h/s (0.832 g) is representative of the maximum possible deceleration of a street car under ideal road conditions. Would it even be feasible to decelerate at that rate if it's icy? The weather conditions make it plausible that it was icy:

[5] January 29, 1996 was a sunny, cold winter day. The temperature was in excess of minus 30. An electronic warning sign for northbound traffic just south of the Calf Robe Bridge had a lighted message which read “Frost warning on bridge deck – reduce speed.”
[72] There was a warning sign telling drivers to reduce their speed. However, there is no evidence that ice contributed to the accident. While drivers that drove after Ms. Eirikson’s accident noted that the bridge was icy, the drivers who came before did not experience ice.

But they disagreed on whether it was actually icy:

[50] Samac Engineering rebutted some of the conclusions in the report by Collision Analysis. Of significance is the comment by Collision Analysis that Ms. Eirikson had begun a lateral shift to the right at a distance sufficiently south of the pumper to avoid a collision. Samac says that the tire marks start on the left side of the laneway and proceeded in a straight line towards the impact location which is indicative of the vehicle braking and sliding because the wheels were locked.
[51] Dr. Nelson provided a rebuttal report to Mr. Navin. He points out that given Mr. Navin’s conclusion that 2,300 to 2,400 vehicles per hour were passing through the northbound lane, that means 30 to 40 vehicles were northbound at the time of the accident. In response to the Collision Analysis’ report, he points out that the first reaction of a driver is try to stop rather than attempt a avoidance steering manoeuver such as trying to drive by the pumper in the median lane. Given the lack of good lane markings, the icy conditions and the emergency occurring on the other side of the road, the average driver would not have been able to satisfactorily process the information available to perform the avoidance manoeuver. To expect a completely rational automated type of response under the circumstances is to misunderstand the biology of the human.
[52] Albert Lund provided a rebuttal report in conjunction with a site visit on May 31, 2000. He drove the route four times between 11 a.m. and 2 p.m. He said that as he was driving the route he still had an uncomfortable feeling as he rounded the long curve approaching the accident location. He felt ‘being locked into his lane’. He described the movement involved in attempting a lane change and said he felt restricted in his movements.
[53] Collision Analysis also provided a rebuttal. They point out that although the tire marks used by Samac to conclude that Ms. Eirikson had braked were likely from her vehicle, they could well have been made by the vehicle that went by after the accident. They say that Ms. Eirikson’s situation was not a complicated one. She was presented with a vehicle that was not similar to most other on the road at that time. It was an emergency vehicle painted in unique colours with flashing warning lights. As well, she had two lanes into which she could move. They say that it is impossible from the photographs to conclude from the tire marks that Ms. Eirikson’s wheels were fully locked thereby rebutting Samac’s conclusion that she could not institute a steering manoeuver. In response to Mr. Lund’s rebuttal report, Collision Analysis seems to be making the argument that many of the perceptual difficulties are ones experienced by Canadian drivers everyday. Finally, they point out that Dr. Nelson made arithmetic errors in Ms. Eirikson’s reaction time.

Finally, is it true that having a large difference in masses (between a fire truck and a car) would make the calculation (conservation of momentum equation, and either conservation of energy or calculation using coefficient of restitution) inaccurate?

[44] [... Collision Analysis] concluded that because of the different masses of the two vehicles, because they could not examine the damage sustained by the two vehicles and because the Eirikson vehicle was involved in the second collision, there could be no accurate prediction of Ms. Eirikson’s pre-impact speed.

 A: You've misinterpreted the quote. The quote describes stopping distance, not stopping time. The distances are expressed as times at a certain starting speed. Context from the quote suggests that this is in order to establish how soon after the threat became visible the driver would have needed to engage the brake in order to come to a complete stop in the time remaining to her. Multiplying the times by the speeds gives the stopping distances: 78m, 94m, and 113m respectively. The actual stopping times and accelerations can be approximated given the distance, initial, and final velocity, given roughly constant acceleration while braking.
Let $v_f = 0$, $\Delta t$ = time spent braking, $\Delta s$ = distance traveled from start of braking to stop of movement
$\bar v = (v_0 + v_f)/2 = v_0/2$
$\Delta t = \Delta s/\bar v$
$a = \frac{\Delta v}{\Delta t} = -\frac{v_0^2}{2\Delta s}$
we have, for $v_0 = 100km/h$,
$a = -4.1 m/s^2$
$\Delta t = 6.7s$
Whether ~$4m/s^2$ is a reasonable expectation or not is a question the manufacturer's specifications might be able to answer.

To the last question, yes. When a stationary vehicle is struck by a moving vehicle in a collision, it can shift forward on its wheels. The distance it moved, cross-referenced with its known braking acceleration, can be used to estimate the collision velocity from conservation of momentum. (Conservation of energy is useless - vehicles are designed deliberately to conserve as little kinetic energy as possible in a collision, spending as much of the collision energy as possible to the deformation of the unoccupied parts of the car's body.) If the stationary vehicle is too much heavier than the vehicle that struck it, however, it's not likely to move an amount that's larger than the margin of error of the measurement, if at all.
