First reason: the ISO sensitive characteristics of film are not exactly comparable to the ISO sensitive characteristics of digital sensors. They are the results of a different set of parameters reading. The two are broadly comparable, but not really exactly comparable. Using a DSLR is sensible if you have not alternatives and will bring you quite in the ballpark, but it's not the "real thing".
Second reason: when you use a DSLR to measure an exposure the camera will actually probably take the picture. Each camera has a certain "mistake" in aperture and shutter speed (real value somehow different from theoretical one). By using a real camera (as in using a DSLR) you are using the entire "system", including the deviances from theoretical values of all elements of that system. Your film camera will have its own deviances, but they will not be superposable to your digital camera. More specifically, if you are using a zoom lens with variable aperture that will greatly influence the reading, you don't easily know the exact "f/value" of a zoom at a certain focal position. The camera meter compensates.
Second reason, variant b: your camera might be using some form of "matrix" metering (or "pattern" metering): you don't know exactly which areas the camera based its measurement on. (That applies to using any SLR as a light meter).
Third reason: the characteristics of the boundaries in shadows and high lights in let's say slide film and digital capture are very different:
Highlights: gradual less-than-proportional fading into white for slides; abrupt burning in digital;
Shadows: gradual less-than-proportional fading into black for slides; increasing levels of noise and decreasing levels of detail (increasingly "muddy" appearance) for digital. Not easy to explain, but the visual effect is different. Not easy to choose equivalent "cut points".
Besides, digital boundaries are not easy to judge even using histograms. Typically histograms are drawn based on in-camera JPEG not on what would be possible with raw files. Moreover, if you have single highlights which are going to be burned (let's say streetlights during a nocturne shots) histograms become useless.
Histograms are used with digital cameras for the "expose to the right" strategy:
That is not applicable to film. The "expose to the right" equivalent with film is reading the highest significant (detail wanted) highlights with a 1° spot meter, and "placing" around 2.33 or 2.5 exposure values above middle grey. (the highest you place it the more you will place it inside the "foot" of the slide film, thus losing detail).
Histograms will not give you that value mainly because you don't know where, in the histogram, is your "highest significant highlight" and where are all the highlights that can be burned (let's say street lamps at night).
If you use negative film, supposing you need to really exploit all the huge dynamic range (an architecture shot of an interior with an intelligible scene outside for instance) the DSLR simulation is going to completely deceive you regarding your real possibilities with negative film because the dynamic range of any digital camera is always much narrower than a colour negative film.
http://www.kenrockwell.com/tech/free-digital-camera.htm (check the bathroom image: even with a fill-in flash, the sky with any digital camera in such an image would certainly be burned white).