My own experience with the diurnal occurrence of Es is very much like the graph that Pat has posted. I find Summer Es is most common around 1500-1900 local solar time (to most of us in the US, about 4:30-8:30 PM), with Es not rare from 0930-1500, and again from 1900-2030.
When it's summer, and you're thinking dinner, you should be thinking skip.
One thing I and other DXers have noticed is that Es that is either off-hours (2300-0700 local time) or off season (September to November or February to April) is far more likely to be from the far north (e.g. Canada) than typical Es.
What is very true is that the distribution can vary greatly from year-to-year (some years there is no morning Es, others, mostly morning Es).
DTV Es with a dipole should be quite possible if there are lowband DTVs in the "sweet circle" (1400-2100 km DX) and the dipole is removed from sources of VHF RFI (e.g., outdoors in a park).
One barrier to DTV Es from greater than 2400 km is that the propagation paths that would support such a great distance often incur more multipath and fading - both of which will thwart a DTV decode. VHF high DTV Es, on the other hand, is only unprecedented because of the sheer rarity of Es above 174 MHz itself. I predict that when there is a good Es opening on those frequencies, more DXers will log a highband DTV than would have logged an analog TV by highband Es in the past - thanks to PSIP.
Es from distances less than 1300 km are far more common on CB, 10m, VHF-low mobile and 6m than they are on TV or FM, not only because of the increased bending with lower frequency, but because the simpler antennas used in these services send more RF (signal) above the horizon and into the ionosphere than the antennas used by TV and FM broadcasters, who focus the RF at the horizon for maximum local coverage.
There is a distinct difference in the behavior of Continental tropospheric propagation (i.e., over land) and Marine tropospheric propagation (i.e., over water, which includes large freshwater lakes, as well as seas). Continental tropo is best in the late summer, and rare in the afternoon. It tends to improve gradually after sunset, and is best from midnight to about 2 hours after sunrise, but may continue into the morning (I have heard many reports from DXers of very long DX around noon, or 1030 solar time, just before the tropo dies out). Both off-hour and off-season Continental tropo are far more common than off-hour/season Es or Marine tropo.
Continental tropo is hampered by local terrain far more than Es.
Continental tropo often (but not always) affects lowband TV and FM broadcast, as well as highband and UHF.
Marine tropo is best when warm, humid air gets pushed over cool water. It is best in the spring (Gulf of Mexico) or the late spring and early summer (Atlantic Ocean and Great Lakes), and is best from afternoon to mid-evening (1200-2100 local solar time). It can often be exploited with very modest antennas very close to the surface.
Under usual Marine tropo conditions, conditions deteriorate significantly with any distance from the beach. I found tropo DX to be far better on the Lake Michigan beach at Manistique than it was at a cabin on Indian Lake (only about 5km inland over flat terrain from Lake Michigan)
In my own observations, Tropo over Lake Michigan had little or no effect on lowband VHF.
While it may be noted that Marine and Continental tropo usually occur at different times, co-incidence of the two modes has happened on occasion. Fernando Garcia in Nuevo Leon, Jim Pizzi in New Mexico, Pat Dyer in Texas and John Combs on Florida are DXers who were successful with very-long-haul "hybrid" tropo DX.
All tropo is severely hampered, if not obliterated, by any significant wind.
Comparing Sporadic-E skip to skip on the AM and shortwave bands is like comparing apples and oranges.
Comparing tropo to skip is like comparing apples and bacon cheeseburgers.