Preparatory plates are silica gel on the front of a glass plate.
If the Rf's are close together then one just needs to run the plate longer...of course there is a limit as you could just run your product right of the end of the plate. I must sheepishly admit to having done this in grad school! Started my plate and then began talking to a friend...completely forgot my plate. It was a sad day as this was a nucleoside compound that I had been working on the synthesis of for well over a year at that point. But I digress...
You probably can't separate similar Rf's on a regular TLC strip, but a plate may very well give you enough silica to do the job.
Even if you get spots that bleed together you can use an exacto knife and cut the main region of the two spots out which would yield pretty pure product.
In any event this is not going to be an economical method. Plates are pretty darn expensive.
Here's on source with some 20cm X 20cm plates that I'm sure would do the job...well I'm pretty sure they would. For sure I would be using regular TLC and a very small amount of material to make sure the spots at least start separating before committing a lot of material to this method.
There must be a cheaper source out there...this was just a one off google search here as I haven't ordered any plates in years now and was just curious as to a rough price range. IF one could find a source were the cost per plate was $2-$3 bucks this would start becoming an attractive method I think.
Admittedly I haven't done a bit of synthetic work in quite few years now...so if other bench chemists have tried this already perhaps they can point out any potential mishaps in addition to what Endlessness noted?
For what it's worth this method is used to separate close analogs in a Med Chem setting all the time...so I'd be pretty surprised if it wasn't possible.