When my computer is in the same place, the RSS (received signal strength) value it receives from my router tends to stay constant at around -60dBM. However, my bandwidth tends to fluctuate a lot over time. Is there an upper or lower bound to bandwidth given a current RSS?
There's no way to say "RSS -60dBm equals transmit rate X Mb/s" or something similar. WiFi performance depends on many factors.
Bandwidth itself can be a bit confusing context. A wireless standard has a max transmit rate, for example with 802.11g it's 54Mb/s. However this is achievable only in a carefully controlled lab environment. In real world you must calculate in protocol and application overheads, effect of half duplex medium etc. If you get 30Mb/s with g you're doing good. Distance from antenna affects your transmit rate, but not signal fluctuation with a stationary client.
We're talking about radio traffic. There's all kinds of WiFi and non-WiFi interference coming from multitude of devices. This cannot be avoided.
Another point is how do you observe the fluctuation. Are you running sequential tests with speedtest.net, or just looking how many bars Windows WiFi icon shows? With Internet-based speedtests you're not only dealing with your own WiFi, you're also depending on your ISP line and a bunch of other Internet paths. Any disturbance on the way affects your transmit.
Since your device is stationary, we're most likely talking about interference and possibly poor antenna placement. I'd suggest that you download a WiFi scanner application and check your wireless environment. Examples of such applications are inSSIDer (google for v3, that's freeware) or ViStumbler for Windows. It's also available for Android in Google Play Store (free) along with a bunch of similar apps. For OSX you could use for example NetSpot.
Most of them will show output similar to this:
This shows you how heavily used each channel is (in this example on b/g 2.4Ghz spectrum). Set your router's radio to channel with least usage. If you're on b/g, in Americas you should choose 1, 6 or 11; elsewhere you can also choose 1, 7, 13. This because the channels overlap, so if you set your radio to for example ch3 you're getting interference from neighboring radios on both ch1 and ch6.
BTW while scanning don't associate your device to the WiFi, that'll skew the scan. Note that these applications will only show you WiFi-related info, not non-WiFi interference.
Try moving the router to a different place. Under the table, in a cupboard or behind a TV are not ideal locations. Keeping the antenna(s) upright is (usually) the best option, as is having basically line-of-sight between the antenna and your device. And don't hang anything on the antenna(s) or place the device right next to a metal plate.
Yes, sometimes different standard can give you a better performance. If your devices are capable of n on 5GHz spectrum, test setting the radio on that. 5GHz has much less usage (and interference) than 2.4GHz. If you're currently on n, test setting the radio to b/g. Note that you only configure this on the router, not on the client.
Test with different driver versions - sometimes older may work better. Also I'd suggest that you check which Wireless NIC your device has and google if other people are experiencing issues with it.
The performance envelope
Yes there is an upper and lower bound given a current RSS. The bandwidth available is determined by something called the modulation coding scheme (MCS). The radio in your endpoint will have an RSSI to MCS mapping that determines the likely successful coding scheme for the current RSSI i.e. the modulation, coding rate (error correction) and spatial streams required to achieve a certain bitrate without constantly dropping out. You see this in action as your connection speed drops when increasing the distance between yourself and the wireless AP. Below is a table showing you some of the factors that contribute to the achievable data (PHY) rates:
Application performance depends on the 802.11 protocol that you are using. Funnily enough, at certain RSSI levels, 802.11g and 802.11n is comparable in performance. 802.11n experiences a higher initial loss from bad RSSI than 802.11g. Without investigating further, my hunch is that this is probably due to the complexity of the modulation in 802.11n.
What does this mean?
In some locations with poor RSSI, 802.11g may actually lead to a better experience with certain applications. The performance difference between 802.11g and 802.11n when using VoIP is usually negligible due to the lower data rates required. For applications requiring high-speed data transfer you really can't go past newer technologies such as 802.11n/ac though.