TLDR: Tropical storm watches/warnings likely for Bermuda today, Lee will drop off the key Friday there on the way to Nova Scotia (with wind/rain for New England). Mashup of AL97 and AL98 might do something or not – will take a few days to sort out.
Here are NHC’s Key Messages regarding Hurricane Lee (en Español: Mensajes Claves), and my TAOS/TC impact estimate based on the official track forecast:

On this track and intensity (and size) forecast, Bermuda will experience tropical storm conditions – 60mph winds, which generally causes minor damage, power outages, that sort of thing that they’ve had this year from other near misses. After than, the longer term forecast shows a landfall somewhere in the Canadian Atlantic Provinces in about five days, although landfall could be in Maine or even as far south as Cape Cod although that’s unlikely. Either way, two things will be working against one another: the storm will be weakening, and will likely be below hurricane force at that point. On the other hand, Lee’s wind field will be expanding and changing, resulting in a wider swath.
On thing about Lee that NHC has been highlighting are the waves. The way that Lee is moving is conducive to creating heavy swell along the coast, and that means rip currents. Here’s a NOAA safety poster for some tips:

It’s likely that the mashup of two disturbances in the far east Atlantic will be causing the usual social media and TV weather folks to soil themselves in excitement. AL97 is in the process of being absorbed by AL98, and once that mashup moves west it has some potential for development. The 00Z GFS run wasn’t able to get a lock on the track as an organized system, but ECMWRF carried it along a track similar to Lee. However, and this will get the scary weather profiteers fired up, GFS shows the system spinning up into a monster storm and stalling right off the coast of the southeast US coast between 240 and 384 hours (10 to 15 days). What are the chances of that actually happening? It’s way too far out to even speculate on that. By this weekend things will perhaps be clearer – in the meantime, don’t get excited about long range model forecasts. Lots can (and will!) happen between now and then, and long range (beyond 5 days) models for storms that don’t exist yet are so uncertain that it’s not worth wasting time talking about (even though I just did 😛 ).
Update: To show you why you shouldn’t get excited about these runs, let’s compare the 00Z and 06Z GFS forecasts for Friday 22 September at 00Z (that’s Thursday evening at 8pm Eastern Time. Here is the ZOMG 00Z run …

and here is the totally “meh” 06Z run for the same forecast time:

As I constantly rant, don’t get excited until there is run-to-run consistency.
Thank You!