I started a thread like this once, asking for more definitive numbers to "shoot for" as opposed to an arbitrary % on a slider (100% of 10 != 20% of 200, for example), but the general response I got was "it depends, are you keeping up with the AIs, and able to broker your own research to get more bang for your beakers through trades".
Once you start to get into the higher levels, trading becomes a key component of research. Lets face it, the AIs have some pretty predictable teching patterns. By using your beakers on "offside" lines that still accomplish a strong goal, you can effectively double or triple your beaker "output". Aesthetics (and then Lit) is a good example of this. The AIs tend to delay that line quite a bit, which allows the Human player to tech it, achieve a solid goal (building the Great Library), and then broker those techs for the ones they skipped. If you dont build the Shwedagon Paya, the AI will put a farily high value on Aesthetics.
So the answer to your question is "It depends". If you are teching at a reasonable pace with the AI, and able to trade your way to equality every now and then, then you have enough beakers. If you find yourself falling well behind, and unable to even research a decent trade-bait tech, then you need to look for more beakers, be it quicker libraries and Scientist specialists, or more commerce from tiles to convert to beakers.
Sucky answer, isnt it, LOL, thats what I thought when I made a thread like this, but as you play, you will find out how true it really is. Here are some tips to helping you maintain tech parity with the AI, especially at Monarch (my preferred level):
Build the Oracle. Yes, it costs time and hammers, but generally you can take a fairly expensive tech from it that the AI places a high value on. CoL, MC, and Monarchy are all fairly prized by the AI. Shoot for one of them, and you effectively get a multiple of techs from this single Wonder.
Build Libraries ASAP in high-food cities, and run a scientist or 2. Even if you have no intention of getting a GP in that city, you can still help your cause quite a bit early in the game with a bunch of scientists. There is a delay in cottages where they kind of suck, but still need to be worked if possible. Scientists can help you get over this gap.
Trade it once, trade it around. This is pretty key. When you look at the tech window in the Diplomacy screen, see which tech that a bunch of AIs dont have, and trade it to a bunch of them in a single turn. This prevents you from having the "trickle down" effect of trading a tech to 1 AI, and having him trade it to all his buddies, thus diminishing the value of it to you. If your going to trade CoL, your better off trading it to a few AIs for backfill, even if the deals are bad, because otherwise, the AI you traded it to will be stronger from his own trading.
For actual, hard, target numbers, I suggest you look through the various RPC and series games, and get an idea of where the "good" and "poor" lines are. There are series and RPCs being run from all levels, Noble to Deity, and a quick check of screenshots in those reports, and how THEY were doing with those Beakers per Turn, can give you a good idea of the variables involved. Another great place to get strong tech-rate info is reading the Succession Games. They also run from Noble to Deity, and they are generally played at a stronger "level" of intensity, and players try to focus on every speck of micromanagement they can in their turnsets.