Remember those days in grade school where the teacher would put a huge jar of marbles on her desk and ask the class to guess how many were in the jar?

No one, not even the smartest second grader, ever came close to the right answer. And since the teacher never, ever actually counted out the marbles we likely will never even know if the 'correct' answer was correct, now that I think about it.

I got thinking about this today as BetaNews reported on the fact that at least 100 more new handsets were going to hit the market by year's end. And just over a third, about 35, will be 3G phones.

That's in addition to the dozens that have already hit the shelves this year.

According to Matt Thornton, an analyst at Avian Securities, coming to market are the following (just in time for back to school and the holidays):

• HTC's Dream on Google's Android platform
• RIM's Thunder, Kickstart, and Javelin devices
• Sony-Ericsson's Xperia (though reports today state this could be on shakey ground)
• Nokia's 5800 Tube, 7510 Supernovia, N79 and N85 (Nokia just released 3 the other week)
• Samsung's KC910
• Motorola's Alexander, Atila, and Ischia devices (about 35 total from what Motorola stated during its most recent earnings call )

Motorola, in particular, was cited by Thornton as in a "tough battle" and the "most challenged."

Which is nothing new, but it's not clear if Thornton's numbers include Moto's 35 new handsets (the vendor is aiming to have 50 new ones total this year).

And doesn't it say something about Moto if the 'challenged' vendor is able to push out a total of 50 in one year?

The big question, though, is how many handsets is too many? Or rather, can there be too many?

Even a bigger question, at least in my mind, is when will some savvy handset maker go the Dell custom PC route and let people build their own smartphone.

I broached this 'smartphone' by design concept to an industry analyst.

I will admit that at first he didn't get my concept, which I think is pretty simple. Users would log on to a handset maker's site, and just like Dell letyou design your own system we'd be able to design our own smartphone - from form factor to memory size to keyboard choice to applications.

I don't think it's as outrageous an idea as the analyst essentially thinks it is. He pointed to the fact that there are more moving parts to a handset.

I responded that users don't have to have decision-making on every part.

Maybe a modified format could be choosing various features and the handset maker offering up the closest designed options.

Neither the analyst or I have any idea of how fast assembly and delivery in such a product design scenario would work. And obviously there are limitations but I think they could be easily overcome.

Users would get exactly what they want with a phone, and I'd think brand loyalty would grow.

So what's your take? Is it a half-brain idea?

After all how many smartphones can fit in a jar?

How many handsets are needed, I wonder. How many different smartphone variations are there already? Will there come a time when every possible combination of smartphone feature has been created?