Yes, a dipole for 50 MHz is most definitely possible.
Your 6 meter band is 52 ± 2 MHz. 2/52 is just over 0.038. This is the frequency variance involved: how much, in terms of a center frequency, does the frequency vary the most while remaining within the band edges. You could make this worse by cutting the antenna for one band edge and using it at another, but it doesn't get that much worse even then.
For comparison, assume 80 meters is 3.5 to 3.8 MHz. That's 3.65 ± 0.15 MHz, where 0.15/3.65 is just over 0.041. If you were to cut an antenna for 3.7 MHz (good for 80 meters SSB) and use it toward the lower end of the band, this effect becomes even more pronounced.
In terms of frequency variance between the band edges, 80 meters is actually worse, despite being far narrower in absolute terms.
Or consider 23 cm; where I'm at, the allocation there is 1240 through 1300 MHz -- that's a good 60 MHz wide swath of spectrum. (60 MHz isn't even an absurdly wide signal in the microwave range.) However, 30/1270 is less than 0.024, so in terms of frequency variance, this band is actually far narrower than is 80 meters.
The antenna doesn't care about the absolute frequency of your transmission. Ignoring any nearby obstacles that could throw off the tuning, antenna matching depends entirely on the error in terms of wavelengths.