335i vs 335d is a good choice as both are right at the top of their respective classes from a performance perspective and there's only a 60kg weight difference.
I can't find comprehensive data for the 335d, but Autocar have published full figures for the 335i and the latest 330d. The 335d has 282bhp and 427lb/ft (vs 241 & 384 for the 330d), so would be significantly quicker than the figures below.
335i on the left, 330d on the right...
0-30mph................. 2.2 sec...................... 2.2 sec
0-60mph................. 5.4 sec...................... 6 sec
0-100mph............... 12.8 sec..................... 15 sec
30-70mph............... 4.9 sec..................... n/a
0-400m.................. 14 / 104.1 sec/mph..... 14.5 / 98.3 sec/mph
0-1000m................. 25.3 / 133.6 sec/mph.. 26.3 / 129 sec/mph
30-50mph in 3rd/4th. 3.1 / 4....................... 2.8 / 4.4
40-60mph in 4th/5th. 4.1 / 4.8 sec.............. 3.9 / 5.4 sec
50-70mph in 5th....... 5 sec......................... 5.2 sec
Top speed............... 155mph...................... 155 mph
Noise at 70mph........ 69 db........................ 67 db
Test average........... 25.7 mpg................... 39.6 mpg
Test best/worst....... 32.1 / 11.3.................. 47.1 / 18.8
In gear numbers for the 330d are pretty much a match for the 335i (little give, little take) so I'd expect the 335d would be right on the 335i if not a little quicker across the board other than possibly the standing start numbers, where it might be a little behind as it requires more gear changes to each increment.
I actually have the 4/09 Car and Driver here where they
review the 335d, which is brand new to the US. The link to the data panel is on the first page.
It shows the 335d very quick to 30, which takes only 2 seconds.
0-60 is 5.7 seconds, which is almost a full second slower than
the 335is they've tested, which do it in 4.8. (
Coupe link)
0-100 for the 335d is about 14.2; a 335i does it in about 12.1.
But the most telling data point is the trap speed in the quarter mile. This is the best overall indicator of a vehicle's acceleration. At the end of the 1320, a 335d is going
100 mph. A 335i is going at
106 mph. 6 mph may not seem like a whole lot at first, but that is difference between a decently quick car (a 335i) and a properly fast one (say, a 997 Carrera S, which is a 112 mph car)
As speeds increase, a 335d is really not any quicker than a 255hp 330i --which makes sense given they have similar power and the torque advantage of the diesel is negated by it's weight.
I would definitely agree though, that the diesel cars due to their flat broad torque curves, perform very well in the in-gear tests and that is their strength. In a typical low speed passing situation (like the 40-60), a 335d is not far behind the a 335i.
But in an all out speed contest, the 335i walks away from the diesel. If you performed, say, a 100-130 mph in gear test, the 335d would get smoked.
The reason you can see diesel engines with huge torque being outperformed by gasoline engines with less torque is the way the engine performs over the rev range. Look at semi trucks. That new Volvo makes over 2000 lb-ft available at probably 1000 rpm, but only makes 700 hp and makes a terrible performance engine because the redline is at 1800 rpm. And that all comes from horsepower being only another way to write torque, but with engine speed taken into account. So in reality, all the performance you need to know is in the torque graph. Because every car will make the same power figure with a set torque figure at a set rpm.
Horsepower is another way to write torque, but with time -ie how quickly work is being performed-- factored in. Torque by itself doesn't tell you anything about how fast something is happening. Only that it happens.
If you want to talk about performance, then time and how quickly something happens ---how fast that torque gets applied, has to enter the picture.
In reality, there is really no reason to chose one over the other, since one value has to be derived from the other. It's like choosing between getting blotto drunk or getting blotto drunk in 30 mins. I'm just trying to point out that relying on a peak torque figure in isolation, by itself, has limited value.
M